Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 27 October 2020


Abstract and Keywords

Tense and aspect have risen to some prominence within linguistics in recent decades as various theories have taken first the verb, and then the inflectional system associated with it, to be the central component of the clause. This has manifested itself most obviously in syntax and morphology, but the effort to understand the meaning and use of time-related expressions has coincidentally played a significant role in the development of new theories of semantics and pragmatics that, in turn, have prompted further research into tense and aspect. This book focuses on tense and aspect, and contains articles on the basic sub-fields of morphology, syntax, and pragmatics, as well as universals and typology, markedness, diachrony and grammaticalization, language contact, and creole languages.

Keywords: tense, aspect, syntax, morphology, pragmatics, universals, typology, markedness, diachrony, grammaticalization

Tense and aspect have risen to some prominence within linguistics in recent decades as various theories have taken first the verb and then the inflectional system associated with it to be the central component of the clause. This has manifested itself most obviously in syntax and morphology, but the effort to understand the meaning and use of time-related expressions has coincidentally played a significant role in the development of new theories of semantics and pragmatics, and those theories, in turn, have prompted further research into tense and aspect. Early in the second decade of the twenty-first century, we can claim to know a great deal more about both subjects than we did when Comrie published his classic works Aspect (1976) and Tense (1985). But as is usual in scholarship, there remain many unanswered questions.


Linguistics is the only field of scholarship that takes language as its primary object of study. But that does not mean that linguistics is the only scholarly discipline that seriously concerns itself with language. Surprisingly many significant advances in the understanding of language have come from logicians and philosophers like J. L. Austin, Donald Davidson, Gottlob Frege, H. Paul Grice, Richard Montague, Terence Parsons, C. S. Peirce, Arthur Prior, Willard Quine, Hans Reichenbach, Bertrand Russell, John Searle, P. F. Strawson, Zeno Vendler, and Ludwig Wittgenstein, all names that will be quite familiar to most linguists.

Where the study of time-related linguistic phenomena—notably tense and aspect—is concerned, in addition to philosophy, literary scholarship, especially in (p. 4) its sub-fields of narratology and stylistics (i.e., literary linguistics), has especially provided a context for that study. While these two fields concern themselves with language, to be sure, ultimately their concerns are divorced from those of linguistics. Nonetheless, both areas have borrowed heavily from linguistics its terms, concepts, data, and methodologies, while contributing their own to it in return. Where tense and aspect in particular are concerned, the interaction of these fields with linguistics is such that no one can safely ignore their importance to, nor the extent to which they provide alternative contexts for, their study.

Yet a third context is provided by computational linguistics (CL), which is where linguistics meets the formal sciences. From its beginnings some sixty years ago there has been considerable tension between three views of CL: as simply that branch of computer science which concerns the processing of natural language, as the branch of linguistics concerned with developing a formal theory of language, and as a component of a multidisciplinary research program related in some way to cognitive science. Whatever CL may be, it provides not merely a special perspective on tense and aspect, as do the various branches of linguistics such as morphology or primary language acquisition, but a distinct context, drawing to a great extent on quite different initial assumptions, applying unique methodologies, and having a distinctive set of goals. Its effect on the study of tense and aspect has been profound.

The Philosophy of Language (Peter Ludlow)

For many philosophers, the main interest in tense comes from the question of whether tense is projected onto the world by the human conception of time, or whether tense in language reflects an objective reality. Modern physics suggests that time is something quite different from our conception of it; in particular, physics provides no support for the notion of tense. Moreover, tense gives rise to a paradox crucial to McTaggart’s argument for the irreality of time. McTaggart (1908) distinguishes the A-series of positions in time, namely the ones that are past, present, and future, and the B-series, which are those that are simply earlier-than and later-than. He points out that time requires change, which the B-series alone cannot supply. However, while the A-series does provide change, that change is paradoxical. Every eventuality (event, action, or state) changes from future to past via the present, but it is a contradiction to say of some particular eventuality that it is (at once) future, present, and past. We recognize that some eventualities were or are future, some were, are, or will be present, and some are or will be past. But here is the paradox, for we must assume the reality of (tensed) time to make such statements, and consequently to prove that there is time.

Notwithstanding this paradox, many philosophers remain “tensers”: believers in the objective reality of tense. The first four sections of Ludlow’s chapter review the debate between them and “detensers,” who do not. In section 4, he discusses the related controversy involving “presentism,” the view that only the present is real. He (p. 5) concludes that the debate between tensers and detensers bears on the metaphysics of time (in the form of presentism) and the philosophy of language (where indexicals, which include tensed verbs, are concerned), and may even have consequences for the question of whether verbal forms are compositional (cf. Verkuyl, this volume), or formed by analogy.

Philosophical debates may seem very far from the concerns of linguists. However, they are not, for formal linguistics (morphology, syntax, and semantics alike) as well as pragmatics depends to a considerable extent on the work of logicians and philosophers. Basic concepts such as point and interval (of time), boundary, temporal ordering, etc., which are part and parcel of the linguists’ toolkit in dealing with tense and aspect, obtain their meaning in the first instance in the context of philosophy, and specifically in the philosophy of language and the philosophy of time. And much as philosophy has contributed basic temporal concepts to linguistics, linguists, through their analyses of actual language and the resultant theories, contribute in turn to philosophy. It should occasion no surprise to see the names of linguists among Ludlow’s references. Some of them have even on occasion been claimed for the philosophers.

Those interested in further pursuing these issues might wish to consult the collections by Le Poidevin and MacBeath (1993), Jokić and Smith (2003), and Mani, Pustejovsky, and Gaizauskas (2005).

Narratology and Literary Linguistics (Monika Fludernik)

In recent times, an increasingly important segment of literary studies has concerned itself with the structures of texts and discourses, from the level of the sentence right up to that of an entire production such as a novel, and thus with the structures and functions of units belonging to different genres, or as Smith (2003, p. 1) says, with “pragmatic factors [in discourse] such as genre expectations, discourse coherence relations, and inference.”

It has become more and more clear that language, and specifically time-related language, plays a crucial role in discourse and text, performing such functions as maintaining coherence; indicating perspective; establishing, maintaining and changing thematic lines; distinguishing foregrounded and backgrounded material; providing narrative movement; marking rhetorical relations; etc. As Fludernik points out (§1), the importance of tense to narrative comes from the significance of time, the fact that “[T]ime is a constitutive component of narrative both on the level of story and that of discourse; time, or the progression of time, axiomatically defines narrative in many definitions of narrativity, i.e., that which constitutes a narrative.” Time—and hence tense-aspect—plays an equally significant role in other genres. To a great extent, time-related language and the structure of discourse (and of text) are intertwined and they can be—and are—studied from either point of view.

Hence an important context for the study of tense and aspect is that of literary scholarship, in particular narratology—the branch of literary scholarship concerned (p. 6) with the analysis of narratives—and literary linguistics or stylistics in the broadest sense. Literary scholars such as Käte Hamburger (1957/1973) and Harald Weinrich (1964/1985), as well as linguists investigating literature and oral discourse, such as Wolfson (1982), Engel (1990), Fleischman (1990), Fludernik (1993), and Carruthers (2005), have made valuable contributions to the study of tense, aspect, and mood.

Fludernik discusses (§2) the pioneering work of Benveniste (1959/1966) and Weinrich (1964/1985) in distinguishing genres of discours and those of histoire (Benveniste’s terms; Weinrich’s are, respectively, Besprechen ‘commentary’ or ‘discourse’, and Erzählen ‘narration’) and their correlations with different systems of tenses: histoire and the diegetic genres with “past” tenses (the preterite, pluperfect, and imperfect), discours with “present” tenses (the present, present perfect, and future). Discours also involves deixis—there is a speaker and an addressee with whom (s)he is communicating in real time, in a dynamic, progressive “now” at a specific locus, “here.” Histoire is free of any deictic centre, of any “here and now,” of any first person who is necessarily part of what is narrated. She focuses on narration to illustrate the analysis of the use and interpretation of tense and aspect in discourse and text, discussing the largely atemporal role of the preterite tense in narration (§3), grounding and perspectival uses of temporal shifting in narration (§4), and chronology (§§5, 6).

Fludernik’s chapter is complemented by that of Carruthers (see below, in the part on Perspectives).

Computational Linguistics (Mark Steedman)

Given the frequent reference in text and speech to temporal phenomena, properties, and relations, the long history within computational linguistics (CL) of tense-aspect-mood (TAM) studies is not surprising. As Steedman points out, theories of temporality have helped develop many computational applications, while computational theory has served to produce theories of the temporal semantics of natural language. At the same time, issues of computability have a long history in theoretical linguistics, going back at least as far as Chomsky’s works of the mid-1950s, and feature prominently in many theories discussed in this volume, notably Relevance Theory and Segmented Discourse Representation Theory. Section 2 of the chapter concerns contributions of theoretical linguistics to CL and section 3, contributions of CL to linguistic temporal semantics.

Steedman discusses (§2.1) the reasons for the relatively uncommon use of temporal semantics and temporal logic in the development of applications such as information-retrieval and question-answering systems, principally the difficulty of separating these from the use of speakers’ knowledge, and the great difficulty of achieving mechanical replication of the human’s apparently effortless formation of associative inferences. General-purpose systems based on linguistic theories are as yet impractical.

Section 2.2 concerns problems of temporal reference. Steedman observes that for “suitably closed domains” it may be possible to “finesse explicit reference” through (p. 7) human contributions such as text annotation and specific rules. But doubt regarding the practicality of human-labeled data has prompted considerable research into unsupervised methods for training learning systems. He concludes that the extremely difficult problem of automatically identifying temporal semantics and reference receives as yet only a partial solution from linguistic semantics, which while crucial to an ultimate solution, is, in itself, insufficient to achieve one.

Steedman points out (§3) that most of the semantic theories discussed in the present volume assume, implicitly or explicitly, a finite but extendable set of rules in human cognition for describing the events that transform one state into another. In one way or another, these theories address the question of the precise content of the states, and the nature of the transformative events.

Models developed for the limited domains considered by theoretical linguistics, when scaled to practical problems involving realistic worlds, become extremely complex. Thus, Steedman argues, theories need to be judged not only in terms of their soundness in representing temporal knowledge, but also of their efficiency for the purpose of searching for plans, asserting that it is because issues of constructivity (the possibility of achieving an algorithm to attain a given state) and efficiency (the possibility of finding proofs with “affordable” resources) are paramount in computer science that theoretical computer science has been the main engine driving progress in the use of temporal logics and that computer scientists have made very important observations about logics of change as they apply to programs and human reasoning alike.

Perhaps this is the greatest significance of computational linguistics for theories claiming to capture human linguistic competence: its insistence on realism, i.e., constructivity and efficiency.


Almost every area of linguistics, with the exceptions of phonetics and phonology, has its own approach to tense and aspect. Not only do morphology, syntax, semantics, and pragmatics differ in their terminology and their methodology, but each area has its own distinct Problematik—they naturally seek to answer quite different questions where tense and aspect are concerned.

In addition to chapters on the basic sub-fields of morphology, syntax, and pragmatics, this part of the Handbook contains chapters on other ways of looking at the problems raised by tense and aspect. The chapters on universals and typology, markedness, and discourse and text complement the ones on the sub-fields mentioned above, while expanding, deepening, or placing within a wider context many of the issues that occur in discussions in those earlier chapters. Here, too, are three chapters offering diachronic (Nicolle on diachrony and grammaticalization), and both diachronic and sociolinguistic perspectives (Friedman on language contact, (p. 8) and Winford on creole languages), as well as chapters on both primary and secondary language acquisition, and on translation.

The first two chapters in this part concern grammatical marking. Grammar is where form meets function, the signified its signifier, and meanings their markers. As Deo notes, for a number of reasons both synchronic and diachronic it is difficult if not impossible to discuss morphological marking and syntactic marking separately. But as the reader will note after a comparison of this chapter and the following one on syntax, morphologists and syntacticians naturally tend to ask somewhat different questions where the form/function interface is concerned.

Universals and Typology (Jean-Pierre Desclés and Zlatka Guentchéva)

In one way or another, most of the chapters in this volume inherently concern universals and/or typology.

Desclés and Guentchéva point out that there are two types of approaches to typology and universals. While many studies, both typological and universal, are inductive and empirical, basing themselves on surveys of a broad and diverse range of languages, others are conceptually based. Chomskyan linguists hold that certain concepts are universal because they are built into the human language faculty. Such “universals” differ from those of the Greenbergian variety (Greenberg, 1963), whether these be absolute or implicational. In section 2 of their chapter, Desclés and Guentchéva distinguish inductive generalizations from invariants resulting from abductive analysis.

Section 3 concerns the typology of aspect, contrasting the largely inductive approach to grammatical aspect, with its morphological focus, concern with grammaticalization paths, and European-style linguistics, with the concept-based approach to Aktionsart, largely carried out within Anglo-Saxon approaches, following Vendler’s (1957) schemata. The authors conclude (§5) that though aspect and Aktionsart enter into a compositional relationship, they draw on a network of basic aspectual concepts (state, event, process, …) and a set of semantic primitives.

Desclés and Guentchéva place the notions of topological interval and (open and closed) boundary at the centre of the conceptualization of aspectuality (§§3, 4), particularly as concerns the interaction of grammatical aspect and Aktionsart, and see boundedness as serving to contrast “complete” situations with “completed” ones, a distinction grammaticalized in some languages, to the perplexity of theorists unable to cope with a category such as the Bulgarian imperfective aorist. They also explore (§4) the relationships and shared properties between states and processes on the one hand and events and processes on the other. This leads to a consideration of resultative states (§4.5), and of the relationship between the perfective and perfectivity (§4.6).

Aspect is highly sensitive to the kind of discourse structures and functions discussed by Fludernik and Carruthers in their chapters, as Desclés and Guentchéva (p. 9) note (§6). They conclude that “a more elaborate and refined characterization of linguistic markers cross-linguistically demands taking into account other factors, generally called discourse factors …, but which in fact also partake in creating meaning and determine how … contextualized markers function,” pointing in the direction, among others, of the issues in the chapters here on case (Richardson) and voice (Vulchanova). In addition, Desclés and Guentchéva distinguish temporal frames of reference (actualized, hypothetical, potential, conditional, irrealis, counterfactual situations, …), thereby touching on matters discussed in the chapters by De Haan and Depraetere.

Morphology (Ashwini Deo)

This chapter bears on many matters discussed elsewhere in this volume, but the emphasis here is on the systems of markers representing the encoded distinctions. Deo “lays out the essential components of a morphologically grounded theory of tense and aspect” (§2): their morphology is about (1) what the possible meanings and the possible markers are, (2) how the two interface, and (3) what we can say about all this from either a universal or a typological point of view. For although languages differ in the formal devices available to them and the systematic contrasts marked, there are important universals and cross-linguistic similarities (§2).

Deo sees a universal theory of temporal/aspectual grammatical meaning as comprising hypotheses regarding the “semantic ingredients” underlying tense/aspect categories, an account of how the temporal/aspectual “pie” can be “cut up” in different languages, and a theory of “the role of defaults and blocking mechanisms in the structuring of tense/aspect systems” (§2). Clearly the construction of such a universal theory is not going to be a trivial task.

Sections 3 and 4 discuss, respectively, tense and aspect and their modes of expression; section 3 principally discusses asymmetry in tense systems, characterizing the present as a default, sometimes morphologically unmarked, tense, often joined with the past or with the future in binary tense systems. There, Deo also discusses the relationship between future tense and modalities, remoteness systems, and portmanteau markers melding tense with other categories.

Section 4.1 discusses the contrast of perfective and imperfective, and, in section 4.2, the progressive and the perfect, while section 4.1 includes a discussion of the relationship between telicity and perfectivity, and contains a survey of some of the theories of aspect regarding this and related issues (thereby touching on matters discussed in the chapters by Filip, de Swart, and Gvozdanović). Though the progressive (cf. the chapter by Mair) is generally classed as a sub-sub-type of the imperfective, it is interesting to learn that encoding of the progressive is independent of that of the imperfective (§4.2). Also, the range of readings available to the perfect (cf. the chapter by Ritz) in a given language is shown to be in part a function of how it is marked.

Section 5 concerns the consequences of the relatively greater semantic scope of tense over aspect for aspects of morphological and syntactic structuring; cross-linguistic (p. 10) surveys reveal that the relative order of the markers of tense and aspect do tend to reflect their relative scope. This section also discusses the ways in which synchrony is contingent on grammaticalization.

Section 6 surveys the results of the typological/grammaticalization tradition founded by Comrie, Bybee, and Dahl, concluding that there is a strong correlation between the meanings of tense-aspect categories and their formal expression, that the sources of the lexemes used to create new tense/aspect morphology lie in a limited number of semantic fields, and that the exponents of categories tend to evolve to express other categories (cf. the chapter by Nicolle).

What Deo says of the theories surveyed earlier in her chapter is a fair representation of the situation for tense/aspect studies as a whole: “The morphological findings from the grammaticalization literature present a challenging explanandum for theories of tense/aspect meaning….” For as we learn more about the how and the what of the evolution of tense/aspect systems, we will be challenged to account for the why.

Syntax (Tim Stowell)

As Stowell says (§1), syntax is a large topic, and so he chooses to focus on just four questions:

  1. 1. What counts as a tense, from a semantic perspective, and from a morpho-syntactic perspective?

  2. 2. Where do tense morphemes occur in syntactic structures, and what role do they play in syntactic derivations?

  3. 3. To what extent are the semantic properties of tenses reflected, in whole or in part, in their syntactic form?

  4. 4. What parallels exist between tenses and other types of grammatical categories, such as verbs, pronouns, or adverbs?

A comparison with Deo’s three main questions (above) reveals considerable overlap between the concerns of these two chapters, but naturally there are important differences of viewpoint, not least because Deo is concerned to a great extent with the mapping of semantic categories into their morphemic exponents, while syntax tends to view semantics as a matter of interpreting morpho-syntactic structures, a perspective reflected in Stowell’s first two questions. Nonetheless, the review of Reichenbach’s (1947) and Comrie’s (1985) theories of tenses (in §2) resembles Deo’s chapter in its focus on the third of Stowell’s questions, and the first half of the first. Unsurprisingly, he finds inadequacies in these accounts, especially Reichenbach’s, not dissimilar to those noted in other chapters, notably Verkuyl’s.

As Stowell says, in section 3 he “review[s] some of the major theories about the syntactic phrase structure associated with simple tenses and periphrastic (complex) tense constructions, based mainly on considerations of morphology and syntax rather than semantics.” This section moves from Chomsky’s early accounts (§3.1), (p. 11) based on affix-hopping and a rather flat structure for the verbal complex of the sentence, through a more hierarchical treatment in which auxiliaries head their own phrases (§3.2), to X-bar theory and the theory of categorical distinctive features (§3.3). The theory of affix placement that resulted was incompatible with theories of the semantics, but was replaced (§3.4) by a theory in which the inflection served as the head of an IP constituent (i.e., the sentence, treated as phrasal in structure) and then by a further development in which features served as phrasal heads, achieving a maximally non-flat, hierarchical structure. From Chomsky’s (1957) theory in which affixes moved, we have arrived at an account in which it is the verbs which move, and certain differences between languages involve a difference in the parametric settings regarding the conditions under which those movements take place. This further evolution fostered an attempt, described in section 4, to reconcile syntactic and semantic accounts of tense. Here, Stowell raises some of the issues and recounts some of the arguments relevant to the subject of the chapters by Hatav, and Ogihara and Sharvit.

Markedness (Edna Andrews)

Markedness is often appealed to in linguistic argumentation, and as Andrews observes, it has been widely applied in a number of research areas, including grammar, semantics, pragmatics, psycholinguistics, and applied linguistics, and, although most characteristic of Slavic linguistics, the concept appears in general linguistic work, including, notably, the study of tense and aspect. But, as she further notes, however “commonplace and to be taken for granted” the concept of markedness may be, it remains controversial and its application to specific problems debatable.

The main reason for this is the multiplicity of concepts of “markedness” and different definitions of the term “marked” that obtain in the various sub-fields—especially morphology, grammar, semantics, and language acquisition. Batistella (1990, p. ix) comments that “Different approaches to markedness (and there are many) define the markedness relation in different ways, apply the concept to different domains of inquiry, and integrate it into different theoretical approaches.” For Lyons (1977, p. 305) markedness describes “a number of disparate and independent phenomena.” Jiang and Shao (2006) distinguish three kinds of markedness—formal, distributional, and semantic. Martin Haspelmath’s “Against Markedness” (2006) distinguishes “[t]welve different senses [of the term], related only by family resemblances” and argues that “markedness” is superfluous—that some of the concepts it denotes are not helpful, and as regards the rest, they should be replaced by “more straightforward, less ambiguous” terms.

Given the frequent recourse to such a problematic and controversial concept (cf. the chapters here by Carruthers and Govozdanović, for example) it is useful for Andrews to review (§1) the development of the theory of markedness and its extension beyond the original phonetics and phonology. She critiques (§2) three “myths” (p. 12) that have arisen in the absence of definitive formulations of markedness theory (cf. Andrews, 1990): that markedness is correlated with frequency, the marked member of an opposition being the lower in frequency; that either morphological or semantic oppositions may be neutralized in certain contexts; and that substitution is definitional for the unmarked member of what Kučera (1980) called a hierarchical correlational opposition.

Jakobson’s theory of shifters provided a unitary account of grammatical categories in terms of the speech situation, narrated situation, and their participants. Andrews presents (§3) revisions of Jakobson’s account by van Schoonefeld and Aronson, who “argue in favor of a different distribution of the concept of shifter, where the verbal categories are either completely characterized by different types of deixis, or the placement of mood and status are reversed from shifter to non-shifter.” As a test case, in Slavic aspect is discussed (§4). Andrews concludes that while markedness can provide a useful heuristic in analyses, by itself it cannot provide complete analyses.

Adverbials (Monika Rathert)

The role of temporal adverbials in temporal reference and in the representation of temporal relations is vastly greater than the relatively small literature on the semantics and pragmatics of temporal adverbials would suggest.

There are literally an unlimited number of such expressions, since they include both expressions incorporating the integers (e.g., for an hour; within two days; three days from now; …) and those which are, or incorporate, phrases or clauses (for the next three days; during her nap; while she slept; …). But even the expressions listable in the lexicon are large in number, including both nouns (today, Monday, …), and adverbs (afterwards, beforehand, hitherto, lately, now, often, soon, thereafter, …), as well as idiomatic words and phrases of sometimes quite complex and/or idiosyncratic formation (examples include forever and a day; from here on in; from now on; in the wee small hours of the morning; never in a million years; nowadays; once in a blue moon; since time immemorial; time after time).

Rathert develops (§1) a semantics for the various types of adverbials, of which she lists just four: positional, which locate eventualities in time—these may be anaphoric (afterward), deictic (yesterday), or clock-calendric (on May 1, 1999); quantificational (once, seldom, sometimes); durational (for an hour); and Extended-Now adverbials (ever since). (Adverbials commonly used with generic, habitual, and iterative expressions are discussed in the chapters by Bertinetto and Lenci, and Carlson.) The discussion of subordinate clauses (in §3) does not include a classification of temporal clauses, but it is plausible that they match these types: anaphoric (after the party; after Sam returned home), deictic (now that Tom is a widower), clock-calendric (when it was just past midnight; when it was no longer 1999), Extended-Now (since Tom has been here), durational (while he was sleeping), as well, perhaps, as quantificational (whenever the sky is blue). Apart from complement clauses, relative clauses (p. 13) also enter into marking temporality: who normally buy pies from the bakers down the street. Much less attention has been paid to other types of embedded, subordinate clauses and phrases, as Rathert notes, and more work is clearly indicated.

The syntax, semantics, and pragmatics of temporal adverbials interact with those of tense and aspect in crucial ways and thus are relevant to every chapter of this book. (The interactions of adverbials with aspect are discussed in the chapters by Carruthers, de Swart, Filip, Gvozdanović, Mair, and Richardson.) The discussion in this chapter focuses (in §2) on the role of adverbials in universal and existential readings of the present perfect, principally in German (cf. the chapter by Ritz).

The adverbial functioning of subordinate clause structures raises questions of their temporal interpretation and the discussion here (§3) overlaps that in the chapters by Hatav, and Ogihara and Sharvit.

Temporal adverbials raise many questions, probably only a minority of which have hitherto been paid adequate attention.

Pragmatics (Patrick Caudal)

Like semantics, the younger field of pragmatics is an extremely broad basic area whose concerns perfuse most of these chapters. Its empirical and theoretical boundaries appear still somewhat unclear (though considerably less so than in the previous decade), and the actual scope of pragmatics varies from one author to another, so that in tackling the central issue—how TAM markers are used and interpreted in context—it is necessary, Caudal points out (§1), to be concerned with questions regarding the semantics/pragmatics interface: “(i) what part of the interpretative content ascribed to tense/aspect forms should pertain to pragmatics, as opposed to semantics? And … (ii) how do semantic and pragmatic phenomena interact with one another in the synchrony and diachrony of tense-aspect forms?” In section 1.1, he contrasts theories taking a static, traditional Montagovian truth-conditional view of meaning with dynamic semantic theories postulating a level of representation prior to model-theoretic truth evaluation, opting specifically for the Segmented Discourse Representation Theory (SDRT) version of the latter, on the grounds that it offers principled manners of constraining pragmatic enrichment mechanism (by making them, for example, dependent on or sensitive to linguistic form and content). Adopting Recanati’s (2004) terminology, he also contrasts “literalism,” strictly separating the contributions of semantics and pragmatics to logical form, with “contextualism,” which allows pragmatics to affect logical form via pragmatic enrichment procedures, and opts again for SDRT’s modulated contextualism; in section 3, he argues against a certain form of radical contextualism primarily resorting to speaker meaning-based, linguistic form-independent pragmatic strategies, as represented by Portner’s (2003) analysis of the present perfect.

For Caudal (§1.1), the main interest of pragmatics lies mainly in two areas, namely (i) “context-sensitive uses of tense-aspect forms in synchrony” and (ii) “the conventionalization of these contextual uses (i.e., the role of pragmatics in the (p. 14) grammaticalization and evolution of tense-aspect forms).” As regards the later, his chapter overlaps Nicolle’s.

Caudal sees objective (i) as “theoretically realized through … different means of modelling contextual pragmatic processes affecting aspectual interpretation, from conversational implicatures to discourse structural parameters such as discourse relations.” In section 2, he argues for a continuum between language-independent, general principles of communication (such as Grice’s theory of implicatures) and language-specific, conventionalized usages. Although he sees Gricean principles as playing a role all across this spectrum, he argues that many phenomena are (at present) best captured within the SDRT framework, particularly those driven by or related to discourse structural parameters, or exhibiting some form of conventionalized, non-free uses (he notably shows how the distribution and interpretation of different (semantic) classes of grammatical aspectual forms interact with different kinds of SDRT discourse relations, and can evolve in arbitrary ways).

Caudal proposes that diachronically emergent uses for tenses may be associated with defeasible pragmatic regularities and yet reflect nascent conventions no longer capable of being “triggered by general, unconstrained pragmatic processes.” He argues that this does not exclude “free enrichment” via pragmatic modulation, but the processes involved in this evolution are generally not entirely free and must not be left unconstrained. He cites as “one of the main defects of a free pragmatic enrichment approach (as in contextualist theories) … the danger of over-generation” and suggests that SDRT “offers an appropriate kind of architecture to try and prevent such over-generation,” while “keeping inferences tractable.” To this end, section 4.1 provides an overview of SDRT and section 4.2 a summary of discourse relations, while section 4.3 discusses the interactions of discourse context, adverbial modifiers, and aspectual classes, thereby connecting this chapter with those by Carruthers, Filip, Fludernik, and Rathert, among others.

Discourse and Text (Janice Carruthers)

Semantics has classically been taken to concern the meanings of linguistic expressions—from the smallest meaningful unit, the morpheme, to the largest, the sentence—without regard for context. Pragmatics, in contrast, has traditionally been defined precisely in terms of the interpretation of a linguistic expression in context, the context being as small as the co-text—the text or discourse surrounding the expression—or as large as everything the participants in the speech act believe about both language and the world.

An important setting for the study of tense and aspect is that of literary scholarship, in particular, narratology, and literary linguistics or stylistics in the broadest sense, the subject of Fludernik’s chapter.

The role that the semantic properties of tenses play in the effects they produce in discourse, especially in interaction with other tenses and temporal expressions (p. 15) such as adverbials, is discussed in section 2 of Carruther’s chapter. In two test cases, she shows that the effects in question arise not exclusively from the semantics of the tenses themselves, but from their interactions with other contextual elements. The first test case is the historical present of French (§2.1). Using the concept of markedness (see above on the chapter by Andrews), she shows how “the contextual effects of [the historical present] arise precisely from features that are not shared” between the present tense and the preterite. The second test case is the French narrative imperfect; Carruthers concludes (§2.2) “that the narrative value of the imperfect … comes from the elements in the context that allow the Discourse Relation of Narration to be posited, and not from the tense, which remains imperfective.”

Carruthers discusses (§3) models from Discourse Representation Theory, Relevance Theory, and Segmented Discourse Representation Theory (SDRT), but specifically adopts an SDRT perspective in her analysis of the role of tense in the construction of discourses, as in recent research on the compound past in French and Australian English (see Ritz, this volume).

Section 4 of the chapter concerns focalization and the creation of point of view in different types of discourse. Carruthers examines the use of the imperfect in free indirect discourse, and the role of the present in contemporary French texts. She concludes that the present in particular “is a crucially important element in focalization strategies in discourse, particularly in modern and post-modern literary texts.”

Just as tense and aspect play a significant role in the construction and interpretation of transsentential linguistic structures and consequently in their study, so that Jahn (2005), for example, devotes one of nine sections to tense, many studies of tense and aspect today turn to the analysis of discourse and text to provide crucial arguments for their analyses and hypotheses. In addition to chapter 10, we see this particularly in the chapters here by Caudal, Rathert, Hatav, Ogihara and Sharvit, Gvozdanović, Mair, and Ritz.

Translation (Diana Santos)

As this chapter makes clear, there is an intimate relationship between translation on the one hand and comparative/contrastive linguistics on the other. Despite a very large literature of mainly bi-linguistic studies, there apparently is no general comparative/contrastive linguistics and little relevant theory outside that which is found in the studies of second language acquisition and of typology and universals (cf. Santos’s section 2.2).

Santos says that it should be hardly possible to discuss translation without involving tense and aspect, but that the reverse is also true, that the practice of translation can and does illuminate many of the issues of tense and aspect found in this book.

She argues (in §2.2) that “translations should be among the primary semantic data for all of linguistics.” She notes their authenticity, their independence from (p. 16) linguistic goals, and their representing the interpretation by the translator of the source text. She sees a purely empirical study of translation practice as less biased than accounts based on assumptions about language universals and typological categories, which may essentially be closed systems and hence unverifiable.

Accordingly, Santos advocates an empiricist, structuralist approach to the study of translation, indeed arguing for linguistic relativity à la Benjamin Lee Whorf, in which languages are each in principle sui generis and may even be incommensurate.

In order to facilitate such an empirical study of translation, she has devised the “translation network model” (TNM, described in §3), based on Moens (1987), and sharing with it (a) the use of a network to describe tense and aspect mechanisms, (b) the use of coercion to explain marked uses of grammatical devices, and (c) the modeling of ambiguity as the existence of several possible paths in the network. One important distinctive feature of the TNM is the assumption that each language has its own sets of aspectual devices and of aspectual categories. Another assumption is that, in addition to fully specified categories, languages display vague categories, which may be further specified in context but do not always require specification for the understanding of a particular text. The description of the use in practice of the TNM forms the bulk of the chapter.

Section 7 concerns the effect of genre on the use of TAM markers, thereby linking this chapter to those by Carruthers, Caudal, and Fludernik.

Diachrony and Grammaticalization (Steve Nicolle)

Up to now the chapters have emphasized synchronic linguistics. The next three look at the diachrony of tense and aspect, which has received much less attention in the literature than have synchronic issues. In recent decades, the focus has been on theories that would make tense-aspect-mood systems out to draw fairly directly on innate language faculties, yet it is clear from the chapters by Deo, Nicolle, Winford, and Friedman that synchronic structures are to a considerable extent contingent on their histories; the importance of diachronic processes for understanding synchronic systems is also illustrated by the chapters by Mair and Ritz, and in another way, by Caudal.

Nicolle starts (§1) by posing the basic questions of how tense-aspect systems come about, how particular markers or distinctions arise, and how systems, markers, and distinctions change over time.

While noting the determinant role morphological typology plays in the expression of tense-aspect categories and the causal relationship between its instability and that of individual markers (cf. the chapter by Deo), he points out that the lack of long-term documentation limits our understanding of morphological typology and the evolution of languages through the famous isolating-agglutinative-fusional-isolating cycle.

Most of the chapter is about the origin and evolution of grammatical categories, though these are inseparable from those of grammatical markers: “primary (p. 17) grammaticalization,” the processes developing grammatical markers from lexical material (§§2–4) and “secondary grammaticalization,” the development of further functions for already grammaticalized constructions (§5).

Cross-linguistically, lexical material of similar meaning develops into grammatical markers of similar category (§2; Deo, this volume). The processes of desemanticization (“semantic bleaching”), the loss of specific meaning in favor of extremely broad, abstract meaning, and decategorialization, loss of unambiguous membership in a specific category, involved are illustrated with the development of the Swahili completive marker sha- (§3). The source material and contexts favorable for grammaticalization are discussed in section 4. (Again, cf. the chapter by Deo.)

Section 5 concerns “secondary grammaticalization,” the development of further functions for already grammaticalized constructions, which Nicolle distinguishes from “actualization,” the spread of a grammaticalized construction through a language community, with concomitant increase of its frequency of use and contexts of use. He notes that similar categories in different languages tend to develop similar additional grammatical functions.

When a set of grammaticalizations are linked sequentially, this is called a “grammaticalization chain” or “path” (§6). Grammaticalization chains in theory represent a set of discrete stages, but in reality it is difficult to identify the moment when a construction shifts from one category to another, partly because it may be difficult to identify the boundary between the categories. The problems this poses for most theories, which assume discrete categories, are the subject of section 7, outlining an approach, based on gradient, non-discrete distinctions, which calls into question the absolute distinctiveness not only of lexical and functional categories, but of “tense” and “aspect” themselves. An alternative approach is suggested in which distinctions between categories are retained but associated with distinct syntactic positions. (Regarding syntax, see Stowell’s chapter in this volume.)

Section 8 concerns the effects of language contact, which is the topic of Friedman’s chapter, and relevant to Winford’s chapter as well, especially its section 3.

Language Contact (Victor Friedman)

As stated above, the synchronic state of a language is not purely a function of universals, of inherent features of human language, but neither is it random. Rather, it is to some extent contingent on its diachrony, both internal and external, though the role of language contact in effecting changes in TAM systems can be, and has at times been, overstated.

The extremes of the effects of language contact include language death, but also the creation of new languages, as explored in Winford’s chapter.

The results of long-term contact among a group of languages can be described as a Sprachbund or language union, the classic example of which is found in the Balkans, where a number of Slavic (Bulgarian, Macedonian, etc.) and other Indo-European languages (Albanian, Greek, Romani, Romanian, …) along with the (p. 18) non-Indo-European Turkish, have mutually influenced each other and have evolved a number of characteristically “Balkan” features that they do not share with non-Balkan members of their own families.

The importance of Friedman’s careful and detailed description of the effects of diffusion on the TAM systems of the Balkan languages for the diachronic study of tense and aspect lies in the sociolinguistic principles it illustrates. Where the synchrony of the TAM systems of languages in general is concerned, diachronic and sociolinguistic studies, like studies on language acquisition, can be revealing not only of the structures of particular systems, but of tense-aspect systems in general, since one cannot change what isn’t there, and the state of a system at a given moment can influence the directions in which it does change.

What Friedman observes are, first, the effects of bi-and multi-lingualism, and second, the impact of multiple systems on the ecology of communication, changes in grammatical functions, creating new categories and markers that alter the grammars of the languages where they occur. Diachronic principles are perceived on the level of whole dialects or languages, but at base they are the product of a complex of sociolinguistic, psycholinguistic, and extrinsic factors on the level of the individual speaker. In the Balkans, there are enough languages with a wealth of recorded history to illustrate many such linguistic phenomena, which, taken together, suggest how contact-induced change in tense-aspect systems participates in the identification and analysis of linguistic areas.

Creole Languages (Donald Winford)

Creoles are interesting from many points of view. Their formation is incompatible with the standard Stammbaum model of historical linguistics, and they illustrate radical effects of language contact (cf. Friedman, this volume), challenging widely accepted views of the extent to which a language can change under the influence of another.

Classically they have presented the enigma of apparent (albeit debatable) uniformity across languages that are geographically widely separated and lexically affiliated to different languages. Winford suggests (§4) that their similarities testify to the role of universal principles in their creation while their differences are evidence of the effects of different linguistic inputs and different sub-paths taken by processes of grammaticalization (cf. §3 and the chapter by Nicolle).

Debates have focused on the contrastive roles in their formation of substratal and superstratal languages, of processes of first language and second language acquisition, of Universal Grammar and general cognitive capacities, and of sociolinguistic and psycholinguistic factors. Theories accounting for their similarities and/or differences include Lefebvre’s Relexification Hypothesis (1986, 1993, 1998); Siegel’s “transfer” hypothesis (1999, 2000, 2008); and Winford’s own approach (2006, to appear).

Creoles have had an especially high profile in debates over the nature of linguistic universals and their role in language acquisition and change, and following its (p. 19) proposal in 1976, Bickerton’s now-superceded language bioprogram theory long served as a connection between creole studies and universalist hypotheses (see too Bickerton 1981, 1983, 1984, 1988). The debate regarding Bickerton’s theory leaned heavily on creole TAM systems, which consequently led to a large literature (including Singler, 1990). More recently those systems have played a significant role in debates as to whether creoles constitute a distinct typological class (McWhorter, 1998, 2005) and whether language universals should be conceived of as innate generalizations written into Universal Grammar, or as part of more general cognitive capacities peculiar to human beings (Winford 2010).

According to the bioprogram hypothesis, creoles arise when children apply their innate language capacity in the process of acquisition. Insofar as the study of creoles accordingly has implications for Universal Grammar, it is important not merely for their study that scholars such as Arends (1989, 1996; cf. Selbach et al., 2009), Singler (1993), and Winford have in recent times rejected the innatist explanation of creole-formation in favor of one rooted in theories of contact-induced change (cf. §3.2), which also takes into account the sociohistorical contexts and social ecologies in which creoles emerged (Mufwene, 2001). This has led to more concern with the processes by which tense/aspect categories and systems are created, and with the role of universal cognitive mechanisms in such processes. Winford claims (§2) a general consensus among creolists that creole formation involves a process of natural second language acquisition by speakers with more or less limited access to the languages that supplied their vocabulary. This has led to increased comparison of creole formation and other cases of second language acquisition (cf. Bardovi-Harlig, this volume) and growing dialogue between scholars in the two fields of study (cf. Andersen, 1983; DeGraff, 1999; Lefebvre et al., 2006).

Primary Language Acquisition (Laura Wagner)

As the last two chapters illustrate, it is ultimately impossible to consider language diachrony without concerning oneself with language acquisition, both primary and secondary.

The central fact in primary language acquisition is that, amazingly enough, children do normally acquire language, and specifically, both the forms and the functions belonging to the tense-aspect system(s) of what will be their native tongue(s). The question is how they manage to do it.

Part of the answer to that question involves, of course, just what it is they acquire. Our understanding of tense-aspect systems and of the processes involved in acquiring them consequently have much to contribute to one another, and more adequate accounts of both will have to develop in tandem. This chapter concludes (in §6) with a number of questions requiring further research; other chapters point to related questions regarding TAM systems themselves.

As the present volume reveals, children are faced with a broad range of formal markers (cf. the chapter by Deo), which may or may not include tense markers (p. 20) (cf. Lin), may include markers on other than verbs (Lecarme), or marking by case (Richardson). Puzzling out the semantic categories and their exponents thus poses a great initial challenge for the child, as is the acquisition of the temporal concepts underlying TAM systems—boundedness, telicity, progression, deixis, etc. These are abstractions which, Wagner points out, are not easily observed in the world. And, we might add, must be abstracted from interpretations distorted by the exigencies of pragmatics. All this at a time when the child has myriad other things to learn, and (Wagner also stresses) must draw on immature cognitive resources.

Much of the literature, we are told, focuses on a systematic phenomenon: children (before about the age of two-and-a-half years) under-extend their semantics by restricting their language’s past tense and perfective markers to telic predicates while restricting its present tense and imperfective markers to atelic ones. With age, they normally under-extend their morphology usage much less, and increasingly their performance exhibits the mature independence of tense, grammatical aspect, and Aktionsart. Their temporal under-extensions would seem to reflect general principles of children’s semantic organization and/or their cognitive architecture. Accordingly, the phenomenon potentially illuminates both temporal semantics and the processes of language development.

Across the many diverse approaches, there is widespread agreement that there is something natural about the associations telic/perfective/past and atelic/imperfective/present, but they disagree on whether they reflect cognition or linguistic organization; perhaps the categories define cognitive prototypes, or perhaps the combinations within each association are less marked or involve less semantic coercion than the cross-association ones. In the latter case, children’s adherence to the associations would reflect their implicit appreciation of linguistic defaults.

All approaches agree that children’s under-extensions reflect the easiest semantic combinations to produce and understand. However, the particular sense in which these classes are easier differs across theoretical positions, and are variously attributed to childrens’ not (fully) possessing the required grammatical resources and to their grammars not having a fully articulated temporal structure.

Ultimately, theories in primary acquisition must account for more than just this facet of production; they must explain the full range of phenomena exhibited in the acquisition of production as well as those of comprehension. Wagner provides an overview of what is known about how children acquire Aktionsart or lexical aspect (in §3), grammatical aspect (§4), and tense (§5).

Second Language Acquisition (Kathleen Bardovi-Harlig)

Bardovi-Harlig connects second language (2L) acquisitional studies with those on first language acquisition. Her chapter focuses on the “aspect hypothesis” (§2), which is based on the observation that children learning their first language encode the completion of events (i.e., aspect) before they encode temporal relations between events and the time of speaking (i.e., tense). The aspect hypothesis (AH), (p. 21) “the most widely tested hypothesis in 2L tense-aspect research,” predicts similar acquisitional patterns in 2L acquisition by adults. Work on this hypothesis serves research into both primary and second language acquisition, since the study of adult learners facilitates the study of language development separate from that of cognitive development.

She further reports (§1) that the developmental sequence in L2 acquisition is generally independent of environment, instruction, or even the learner’s language(s) and the target language. Basic rules for marking are underutilized in interlanguage (the form of the L2 language produced by learners, incorporating features of their native language(s)) relative to their use by native speakers in the target language, beginning at a simpler, more prototypical starting point and subsequently spreading throughout the learner’s L2 grammar.

The focus of investigations of L2 morphology has come to be on form-meaning associations within developing L2 temporal systems. The AH has been prominent in form-oriented approaches which begin with tense-aspect morphology and, based on observations of its distribution, seek to understand what the emerging forms mean in the developing L2. An alternative set of meaning-oriented approaches, championed largely by European functionalists, began with temporal concepts and inquired into how developing systems expressed such concepts (e.g., the past). These inquiries captured pre-morphology stages of interlanguage in which lexical items were used by learners to make temporal reference and later stages in which they gradually added tense-aspect morphology to their L2.

Bardovi-Harlig notes that studies documenting the spread of morphology from lexical category to lexical category are few relative to those that document prototypical associations typical of the initial stages of acquisition, and that even fewer studies attempt an account of how such spread occurs.

L2 acquisition researchers have found a relationship between the use of verbal morphology in interlanguage and the grounding of the narrative. The interlanguage discourse hypothesis (§2.2) predicts that “learners use emerging verbal morphology to distinguish foreground from background in narratives” and, based on language universals, predicts that the foreground would be marked by the perfective past. Bardovi-Harlig compares the predictions of the aspect and discourse hypotheses (§2.3) and provides evidence (in §3) for the former.

Her own account (in §3.3) of how the perfective past jumps from goal-oriented telics to open-ended activities attributes it to the narrative function of foregrounded, chronologically sequenced, activities.

Research on the aspect hypothesis has significantly increased interest in tense and aspect among students of second language acquisition. She reports (§4) on several types of research which do not fall within her AH-oriented focus. As in most of the chapters here, Bardovi-Harlig concludes by pointing out a number of areas requiring further research, one being the need for longitudinal studies to facilitate the observation of more stages and produce more nuanced analysis than has been possible with the largely cross-sectional studies to date.

(p. 22) Tense

For decades, the standard theory of tense has been that of Reichenbach (1947), referred to in many of the chapters here. In naїve accounts, a tense marker references a sector of time (past tense refers to past time, etc.) and it is the function of a tense to situate an eventuality—event, process, or state—in the corresponding time: the past tense verb painted locates a particular act of painting in the past. But hardly any languages actually have the predicted three tenses, and while the naїve notion seems at first glance to work for “primary” or “absolute” tenses such as the past, it fails completely where “secondary” or “relative”1 tenses, e.g., the pluperfect or the conditional (e.g., would leave in the next day he would leave for home), are concerned. Before Reichenbach, most analysts treated secondary tenses as implicitly involving two temporal relationships, where the eventuality is located in a time sphere defined not by the present time, but by some other time. Thus the conditional is a future-“in” (relative to)-the past, the future perfect a past-in (relative to)-the future, etc. This scheme is essentially that of Jespersen (1924) and Bull (1960). What Reichenbach did, as Hewson points out in his chapter, was unify the treatment of tenses by defining all tenses in terms of a “middle term,” the reference point, alongside the time S of the speech act (the present) and the time E of the “event.” Thus, the past is not simply defined by the time of the “event” E preceding the present (E<S); its absoluteness comes from the fact that S serves as the reference time R (R=S), so that the past is defined by Reichenbach as E_R=S (i.e., E<R=S).

The naïve conception of tense, for all its inadequacy before the facts, has proven remarkably resilient. Equally resilient, Hewson argues (in §1), is the mistaken identification of a tense marker with a tense, pointing to echoes of this in Reichenbach (1947) and Comrie (1985).

As shown in Table 1 in Verkuyl’s chapter, Reichenbach’s triples have since been reanalyzed as pairs (cf. Johnson, 1981, §2; and Dinsmore, 1982); the past perfect, for example, is now defined not as E_R_S (i.e., E<R<S), but as R_S (R<S), past, and E_R (E<R), anterior. (This analysis is in fact implicit in the names Reichenbach gave the tenses; the traditional “past perfect” is his “anterior past.”) As Johnson (1981) makes explicit, we may identify the R/S relationship with tense and E/R with aspect. Many contemporary scholars, including some in this volume, reject a purely temporal, Reichenbachian analysis in favour of just such a mixed temporal/aspectual one, in which the future perfect, for example, represents simply a combination of future tense and perfect aspect. But as most of the chapters in this part of the book demonstrate, that is still not the full story.

Tense (John Hewson)

Hewson eschews form-based systems, insisting that tense is a system of concepts (§1.1). Whereas Reichenbach observed that coherence in a sentence requires the permanence of the “reference point” R,2 which roughly agrees with traditional notions (p. 23) of temporal accord between clauses and a sequence of tenses rule (cf. Hatav, Ogihara and Sharvit, this volume), Hewson assigns the property of coherence to tense on the conceptual level, not that of markers (§1.2).

In sections 2 and 3, he presents a Guilleaumian analysis of tense, illustrated in section 3.1 by the system of Indo-European tenses, a binary analysis that reminds one somewhat of Verkuyl’s analysis (in this volume) of English. Guillaume’s ideas have been influential in the French-speaking world, but almost completely ignored in the Anglo-Saxon one outside of Canada.3

Guillaume distinguishes the time containing the event (which provides the category of tense)—“Universe Time”—and the time contained in the event (the various representations of which provide the varieties of aspect), “Event Time.” He also distinguishes “Descending Time” and “Ascending Time.” The former reflects the experience in memory of time coming out of the future and going through the present into the past, where now is always descending into the past, and in this “passive” view of objective time, the past is irretrievable. In the latter, our consciousness moves forward, now is always progressing into the future, and this “active” view is of subjective time.4

The core of the theory of tense presented by Hewson he states thus (§3.1):

  1. 1. The unmarked forms in Descending Time represent events as incomplete, in progress, whereas the unmarked forms in Ascending Time represent events as complete.

  2. 2. There are “Type A” languages with two tenses in Descending Time, like Greek and the Slavic languages.

  3. 3. There are “Type B” languages with two tenses in Ascending Time, such as English.

  4. 4. There are also “Type C” languages (e.g., some Semitic ones) with systems that have tenses in both Ascending and Descending Time, where Perfectives and Progressives are still marked forms, but Imperfectives and “Performatives” (the “simple” tenses) may, in certain circumstances, be marked as well.

  5. 5. These are the building blocks for tense systems in the languages of the world.

Section 4 presents a kind of “ontogeny recapitulates phylogeny” characterization of tense, in which the development of more elaborated semantics is correlated with morphological structure, and the progression from aspectual to temporal concepts in primary and secondary language acquisition (cf. the chapters by Wagner and Bardovi-Harlig) is cited as a parallel. Thus, the distinction of “quasi-nominals” (written, writing, (to) write), the subjunctive, and the indicative is seen as providing a set of levels in Guillaumean terms. Thus one can speak of a progression that is both developmental and diachronic.

In section 5, Hewson analyzes three Bantu remoteness systems (cf. on Botne, below), and in section 6, discusses the temporal use of modal forms, and vice versa (cf. de Haan, Depraetere, this volume).

(p. 24) Remoteness Distinctions (Robert Botne)

A problem for Reichenbach’s and similar analyses comes from languages with tense systems marking remoteness distinctions, so-called “metric tense” systems (Carlson, 2009), the subject of Botne’s chapter and section 5 of Hewson’s.

The distinctions between the tenses which most linguists involve themselves with are qualitative and discrete. The past is simply different from the present. But the tenses in remoteness systems differ quantitatively. At a given time of utterance it is quite possible to have a choice of different tenses with which to report an eventuality, the interpretation of which, moreover, is “flexible” (Botne, §9). These systems are characterized by multiple past (and future) tenses, and may involve a large number of grades (up to seven) and/or great complexity in the determination of which tense to use, which is usually based on a number of pragmatic factors. While sometimes the distinction is calendric and discrete, as between the hodiernal (the eventuality occurred earlier today) and the hesternal tenses (the eventuality occurred yesterday), it often is vague, subjective, and not associated with any objective temporal bound.

The differences between tenses in remoteness systems concern the distance in time between E and the deictic centre (the present). But in a neo-Reichenbachian analysis, the “tense” component of a grammatical tense has to do with the relationship of R to S, not E; E relates only to R, in the aspectual component. We might salvage the situation by observing that in these systems R is S, R=S. But then the quantitative past tenses all become in effect present tenses, as do the future tenses. There simply seems to be no principled way of extending Reichenbachian analyses to the tenses of remoteness systems. One of the goals of Hewson’s chapter, indeed, and one of his motivations for presenting a non-Reichenbachian theory, is precisely to arrive at an adequate account of such systems.

Botne’s chapter provides a vastly more comprehensive and vastly more detailed study of remoteness systems in a smaller compass than has hitherto been available. It has long been known that such systems tend to be symmetrical, but that in asymmetrical systems, past tenses exceed future ones. But the factors determining tense use and distinguishing types of systems, and the interactions of those various factors, have been poorly delineated, and Botne’s study points (§9) toward “a more complex organization underlying the distribution of form and meaning” than has generally been recognized. (For some of the complexities, see §§4–7.)

The analysis in his chapter sets out three patterns of organization in which temporal distinctions are typically grounded: “natural cycles, contrasting time units (‘Currently Relevant’ vs ‘Adjoining’) and time scales (days, months, seasons, years) [cf. §1], life (life span, memory) [cf. §2], or epistemic value [cf. §3].” Remoteness, itself, is shown not to be a simple, straightforward concept, but one that involves differences in time scale and temporal domain. In summing up (§9), he proposes that languages may differ in whether they differentiate time units, whether they sub-divide the Currently Relevant temporal domain, or whether they make manifest a distinct remote domain.

(p. 25) Botne sees no discontinuity between a remoteness (multi-tense) system and an apparently non-quantitative, single-tense system like that of English. The latter is just the limiting case, the simplest of metric systems. Such a claim poses a new challenge to orthodox theories of tense.

Compositionality (Henk Verkuyl)

Verkuyl’s interest is not only in arriving at a more adequate account of tense, but in dealing with the problem of compositionality, a generally accepted principle which is, nonetheless, highly controversial (cf. Szabo, 2000, 2007; Dever, 2006; Pagin and Westerstål, 2008, 2010a, 2010b).

The principle of compositionality5 is that the meaning of a linguistic expression e is a function of, and predictable from, the meanings m1, ……, mi of the immediate constituents c1, ……, ci of e, and the mode of composing those constituents into the hyper-constituent e, its internal syntax. A sentence like John loves Mary thus means what it does because of what its individual morphemes mean, but also (very roughly) because English is an SVO language, with all that that implies.

To be sure, language is replete with exceptions to compositionality. But while there are semantically and/or formally complex tenses in languages in which the form/function relationship is purely contingent, such as the periphrastic “perfect” of Catalan (jo vaig cantar “I sang, I have sung”), formally similar to the periphrastic futures of other Romance languages (je vais chanter, vou cantar, “I’m going to sing”), it is assumed that in general, the marking of tenses is compositional. There are some difficulties with this, however, perhaps the most notorious case being that of the much-discussed -te ir(u) construction of Japanese, which is ambiguous between a completive reading similar to the present perfect and a continuative reading similar to the present progressive.6 This construction has defied definitive analysis; it remains controversial whether it is compositional or not, and its very existence calls into question basic concepts involved in compositionality. But the vast majority of temporal (and aspectual) constructions are clearly compositional.

Verkuyl argues (in §2.1) that Reichenbach’s theory is in principle non-compositional, incapable of providing compositional analyses of compositional tenses. One of his arguments involves the fundamental flaw in the system occurring in the Posterior Future will walk analyzed by Reichenbach as S—R—E. Under a strict compositional approach, Reichenbach is forced to produce will will walk for the Posterior Future, the first will being a finite future form (by S—R), the second will an infinite posterior form (by R—E). A simple reduction from will will walk to will walk is logical-semantically impossible from a compositional point of view.

There are many problems with Reichenbach’s theory, discussed in several chapters here (most notably that of Stowell), especially its over- and under- generation of tenses. Verkuyl argues (§2) for a more fundamental, irreparable flaw: the system is ternary, not binary, built on two distinctions rather than the three Verkuyl inherits (p. 26) (§2.2) from Te Winkel’s treatment of Dutch: past/present, “synchronous”/“posterior,” and imperfect/perfect.

The bulk of Verkuyl’s chapter (§3), however, concerns his compositional theory of aspect, based on (1) the compositionality of aspect, with syntax playing a crucial role in determining the aspectual value of a structure; (2) the Plus-Principle, which implies that terminativity (boundedness, telicity) is a marked aspectual compositional property. He illustrates the utility, and consequences, of these principles in discussing the apparent conflict between Slavic and non-Slavic systems (§§3, 4), explaining the differences between Slavic and non-Slavic languages thus: in Germanic languages the semantics involves tense operators, while in a language like Russian the corresponding information is encapsulated in the predication itself and it has to be sorted out by sticking to predicational composition.

The Supercomposé Past Tense (Louis de Saussure and Bertrand Sthioul)

Among the types of tenses which have received relatively little attention in the literature are the surcomposé or supercomposed tenses, formed with two participles, for example, the French passé surcomposé tense (PST) in Jean a eu chanté. Often thought to occur only in French and Occitan, such tenses do occur elsewhere, and Saussure and Sthioul mention varieties of Afrikaans, Flemish, German, Yiddish, and even English (e.g., I’ve had gone through that experience before)7 among Germanic languages; of Catalan, Italian, Rhaeto-Romance, and Spanish among the Romance; the Slavic languages Macedonian and Sorbian; Albanian; and the non-Indo-European Tibetan, Korean, and Basque. Likely the phenomenon is even more widespread.

What lends the supercomposed tenses their cross-linguistic interest is not merely their unusual morphology. If these tenses are compositional, they seemingly require two reference points for their semantic analysis, something impossible in Reichenbach’s and all other theories of tense that assume a single temporal frame or “reference point.” But if they are not compositional, these tenses suggest that they, and therefore at least some verbal forms, are analogical rather than compositional in formation. Consequently, one way or another, they seem to pose a fundamental challenge for at least compositional theories of tense.

In section 1, cross-linguistic semantic and pragmatic peculiarities of surcomposé tenses are described: the “double perfect” of German may substitute for the simple past or the pluperfect, albeit with modal nuances; the passé surcomposé may in some languages signal a remote past, but in others provide a recent, even hodiernal, interpretation; it may also indicate unexpectedness or chance occurrence.

The French tenses, at least, do not seem to be compositional. However, Saussure and Sthioul suggest (§1), the two auxiliaries do allow for an ambiguity: the past auxiliary allows an interpretation in which the eventuality is primarily relevant at a past reference point R, whereas the present auxiliary may place relevance at the (p. 27) speech act time S. Operating within the approach of the “Geneva School” version of Relevance Theory, they attribute (§5) the variety of effects to “contextual accommodations,” to specific outputs of the interpretive procedure encoded by the tense, “obtained under contextual pressure,” that is, generated by pragmatic enrichments concerning the relevance of the eventuality either at the reference point or at the time of speech.

Along the way they consider, and reject, first a temporal account (§2) and then an aspectual account (§3).

Regarding the relationship of the “regional” PST to that of standard French, they conclude (§5) that the former represents simply a particular enrichment of the “regular” tense, “anchoring on the presentness of the auxiliary, just as with a present perfect, even if speakers of standard French reject them as dialectal….”

They further conclude that universally these tenses indicate relevance in the present, and in addition, in contrast to the present perfect, either remoteness (E<R<S) or termination. They speculate that implicatures “about the possible re-occurrence of the considered state of affairs in the present” become grammaticalized in the tense, the past auxiliary describing the eventuality as remote or terminated, the present auxiliary capturing present relevance. Hence the analysis of supercomposite tenses has consequences for both synchronic and diachronic theories of tense, touching as it does on compositionality, grammaticalization, and other central issues.

Bound Tenses (Hatav); Embedded Tenses (Ogihara and Sharvit)

These two chapters concern the interpretation of tenses in context, and in particular, in embedded structures, where tenses often receive interpretations distinct from their nominal semantic values (even unembedded tenses may do so in some contexts—cf. Fludernik, Carruthers, in this volume).

The current theory of embedded tense is largely based on only a handful of languages, principally English, Hebrew, Japanese, and Russian, but the facts remain unclear even where these are concerned. The contextual interpretation of tenses is complicated, and theories differ as to which phenomena are to be explained in syntactic, semantic, or pragmatic terms. Given the controversies in the area, a great deal of further investigation will be required before a widely accepted theory is achieved.

One parameter assumed to distinguish languages is the tense strategy adopted (cf. Hatav, §1). In Sequence-of-Tense (SOT) languages like English, interpretation of an embedded tense is typically “absolute”: the evaluation time is generally the time of speech. But in non-SOT languages such as Hebrew, Japanese, and Russian, the evaluation time shifts to the attitude-holder’s temporal point of view: the interpretation is typically “relative.” Thus English uses the past tense (that Miriam loved him) to report Yoni’s thought, “Miriam loves me,” but Hebrew uses the present tense, (p. 28) since Miriam’s loving Yoni is “simultaneous” with his thought. Ogihara and Sharvit, however, reject the SOT/non-SOT dichotomy as unrevealing, given the mixed behavior of Hebrew. It has likewise been claimed that Russian is not uniformly non-SOT (Kondrashova, 2005; Kubota et al., 2009a, 2009b).

In either type of language, interpretation depends in part on whether the embedded tense is within the scope of an intensional context as in the case of sentences with attitude verbs such as think, allowing tense shifting to occur, or in an extensional context, which precludes it, as in the case of the relative clause (RC) in Hatav’s example, Yoni met the woman who is sobbing in the corner; is sobbing cannot receive a shifted, relative reading. Treating tenses as variables (à la Partee, 1973), and distinguishing free tenses (§2.1) from bound tenses, she ascribes this behavior to free tenses being evaluated at S, and tenses within the scope of an extensional context being unbound, whether embedded or not.

The interpretation of tenses is, however, also subject to general pragmatic principles, and Hatav advocates a pragmatic analysis in terms of Segmented Discourse Representation Theory. (SDRT is outlined in the chapter by Caudal.)

Crucial for any account of embedded tense is the behavior of past-under-past sentences like Hatav’s (17) John thought that Mary was pregnant, which in English are ambiguous between a “simultaneous” reading (in which (17) represents the thought “Mary is pregnant”) and a “shifted” reading, in which his thought was “Mary was pregnant.”

In accounting for this ambiguity, Hatav follows Ogihara and von Stechow in predicating (as do Ogihara and Sharvit) an optional deletion of the lower tense under identity with the matrix tense. When deleted, the time of the now tenseless predicate is interpreted as the same as that of the subjective time of the attitude-holder (John, in our example) obtaining at the time denoted by the matrix predication. When not deleted, the lower tense is bound by the matrix tense, that is, interpreted relative to the subjective time of the attitude holder obtaining at the time the matrix tense represents. The bulk of Hatav’s section 2.2 illustrates the interpretation of such bound relative tenses in Hebrew.

The chapter by Ogihara and Sharvit (O&S) presents a finer-grained consideration of much the same facts as in Hatav’s chapter (cf. their §2).

They see the quite complicated behavior of a present under a past or a future (cf. their table, (22)) as equally important as that of the past-under-past. In Japanese, a simultaneous reading is always available for such an embedded structure. But in English, it is impossible in the case of a present-under-past, and while in Hebrew relative clauses act like English ones, complements of attitude verbs (CAVs) act like those of Japanese. Thus, two central issues concern the causes of the differential effects of tense configurations, and of the differing behaviors of CAVs and RCs.

In section 3, O&S set out to account for the data by examining and contrasting Abusch’s (1993, 1997) theory based on the Upper Limit Constraint (ULC), and Ogihara’s (1995, 1996) own “copy-based” theory (CBT). These two share commonalities, not only with one another, but with Hatav’s chapter, for example, a parameter for whether or not a language has an optional rule deleting matching tenses. Part of the (p. 29) explanation for the data lies in another commonality, a mechanism for de re tense interpretation (§3.2). However, their somewhat different mechanisms lead to different predictions.

The ULC-based theory assumes that any embedded tense has the option of being interpreted de re (cf. §3.2.1). But it also requires that the reference of an embedded tense not be a time that begins after the attitude-holder’s “now.” According to this theory, then, a de re LF (logical form) of past-under-past may support a “simultaneous” reading (as well as a “back-shifted” reading; when the time description happens to be “a month before now,” as in (15))


A month ago, Joseph found out that Mary loves him.

But the theory fails to explain why the Japanese counterpart of (7a), i.e., (5b), lacks a simultaneous reading, and fails to accurately predict that in Hebrew relative clauses, the present does not produce a simultaneous reading under past.


#2005-nen ni Joseph-wa Mary-ga sono-toki zibun-o aisi-te i-ta-to sinzi-te i-ta


lifney alpayim šana, Yosef xašav še Miriam ahava oto az

“two thousand years ago Joseph thought that Miriam loved him [at that time]”

In contrast to the ULC theory, the CBT distinguishes among different types of languages in terms of whether they treat their tenses as pronominal, quantificational, or both (cf. §4.2). O&S note that in certain cases its predictions are the same as those of the ULC theory, but unlike that theory, it fails to predict a simultaneous reading for the Hebrew counterpart of Joseph believed that Mary loved him.

Ogihara and Sharvit conclude (§1) that “both theories are only partially successful and that each of them accounts for a different aspect of [the] variation” between languages, and in section 4 present further data that they use in section 5 to develop a new theory which, “despite many loose ends,” may “serve as a springboard for more cross-linguistic study regarding the behavior of tense in embedded clauses….” However, O&S warn that the theory may predict non-existent types of languages. This would seem to suggest that the search continue for a simpler, more general account.

Tenselessness (Jo-Wang Lin)

Speakers of languages like English think of tense as natural, but some languages have no tense markings: well-known examples include the Chinese languages (such as Mandarin), Yukatec Maya, and Guaraní. Notwithstanding this lack, they are fully capable of expressing temporal distinctions, though they use aspect to do so.

Far from being the norm, tensed languages may well be a minority: DeCaen (1996) claims that tenseless systems are possessed by at least half of the world’s languages. His further claim that “Virtually all non-European systems, with the major exceptions of the Dravidian, Turkic, and Benue-Congo families [and] the Quechua dialects, are deemed ‘tenseless’” would seem to suggest that tense may be an areal feature.

(p. 30) The existence of tenseless languages challenges our understanding of tense. Is tense merely an areal feature resulting from the accidents of language history? Or is there indeed something “natural” about tense, and if so, what?

At this point the question should be raised of just what a tenseless language is. Lin points out that this is less obvious than it might seem. Comrie (1985), as Lin notes, defines tense as the “grammaticalized expression of location in time” (pp. 9–10), though most scholars have replaced “location in time” with the relationship between a reference (Reichenbach, 1947) or “topic” (Klein, 1994) time and an orientation time, by default the speech act time.

The criterion of grammaticalization would seem to exclude purely lexical devices such as adverbs and prefixes (e.g., ex-), though the phenomenon of nominal tense (discussed in Lecarme’s chapter) may require a more finely tuned definition of (purely verbal) tense. As for aspect, Lin presents (§2) several criteria distinguishing it from tense, some drawn from Tonhauser (2006).

Lin argues (§3) for the tenselessness of Chinese. While the usual arguments for this are semantic, Lin utilizes syntactic arguments. This is because in generative grammar the inflectional system of the verb, including tense, has come to be treated as the central component of the clause. Consequently some scholars, such as Sybesma (2007), have proposed that apparently tenseless languages do have tense, but it is marked by a null. It is primarily to counter the null tense hypothesis that Lin argues on a syntactic basis.

Assuming that there are tenseless languages, the question is how they convey the temporal properties and relations that we expect a language to express using tense. Lin’s conclusion (§8) is that “aspectual information, together with topic time resolution determined by an overt temporal adverbial or discourse anaphora, plays a significant role in determining temporal location in a tenseless language.” Although there are hints in Lin’s chapter of how this is done in Chinese, a general theory of temporal expression in tenseless languages remains for the future, not least because of the unanswered question as to why languages have tense, if temporal properties, relations, and objects can all be adequately expressed in its absence. In any case, the serious study of tenselessness is just beginning.

Nominal Tense (Jacqueline Lecarme)

Exotic as supercomposed (surcomposé) tenses and remoteness systems may be, from the viewpoint of traditional European grammar, at least these are tenses morphosyntactically marked on, or in association with, the verb, much as are more familiar tenses. But there are languages in which temporal distinctions may be marked on substantives (nouns, adjectives) other than deverbal ones, as in Lecarme’s examples (1b, 28b) below:



lálém-cha (Halkomelem)



“my future house” (Lecarme, §1, from Wiltschko, 2003)



wanaagsanayd (Somali)



“the good student (f., past)” (Lecarme, 1999)

(p. 31) Such markings may also occur within noun and adjective phrases elsewhere than on the lexical head itself.

For both semantic and formal reasons, Lecarme argues (§§1, 2) that such systems of marking, commonly referred to as “nominal tense,” do in fact represent tense, differing from verbal tense only in the type of expression tense is marked on.

While definitely unusual, nominal tense is more widely spread than one might think. Languages cited in the literature include American (Chamicuro, Guaraní Halkomelem, Kwakiutl, Nootka, St’á’imcets, Tariana), African (Somali), and Australian (Jingulu, Kayardild) ones. The full extent of the phenomenon is unknown and further research will undoubtedly find many more languages in which it occurs.

The existence of tense marking in substantives naturally raises questions (§§1, 2) such as how nominal and verbal tense interact, whether temporal reference works the same way in both types of tense, whether the relevant syntactic structures are similar in the two cases, and what we can, and must, say about the grammaticalization of tense in substantives. And, further, what implications does nominal tense have for theories of tense in general?

Lecarme discusses (§3) Enç’s claim (1981, 1986, 1987) that the temporal interpretation of noun phrases is independent of that of the verb in sentences, concluding that while the temporal location of individuals is not directly determined by the tense of the sentence, the time variables of nominals can receive a value from the context, phrase-internally from temporal modifiers, or from nominal tense and aspect markers. She says that in the framework of Chomsky (2001, 2008), weak (temporally indefinite) NPs are merged in the VP and interpreted in situ, whereas “presuppositional” (tense-related) phrases occupy the positions typically associated with “surface structure” interpretations.

In section 4, she discusses further implications of nominal tense for the syntactic structure of nominals within a Minimalist framework. Lecarme proposes that the temporal structure of nominal phrases is strongly parallel to that of clauses, with a category D formally parallel to a C, and that parallel to the C-T relation expressing clausal finiteness, a D-T relation expresses “nominal finiteness.” Nominal tense seemingly exemplifies the well-known syntactic and semantic parallels that exist between sentences and nominals.

Like verbal tenses, nominal tenses have modal uses (§5) and specifically can convey evidentiality (cf. de Haan, this volume). Lecarme discusses the morphosyntactic and semantic connections between the category of remoteness and that of invisibility and the superimposition in some languages of the visible versus non-visible opposition on a proximal/distal system. She proposes that the common feature underlying the meanings of the past morphology in nominals is an abstract feature of “exclusion/dissociation.”

(p. 32) Aspect

The concept of “aspect” comprehends at least two distinct systems of categories that are nonetheless so intertwined that they have been, and still are, difficult to prise apart. Aspect “proper”—grammatical or verbal aspect, the subject of the chapter by de Swart— is a sub-system belonging to the grammar of a particular language, much as tense is. Aspect in this sense interacts with the non-grammatical, language-independent categorization of types of eventualities and/or their lexical expressions (the subject of the chapter by Filip) variously called lexical aspect, actionality, or Aktionsart “type of action”. For example, illness (i.e., being ill) is a state, and any expression in any language meaning “be ill” is stative, which has implications for its semantics and pragmatics.

No consensus has yet emerged, however, as to the meanings and uses of the terms lexical aspect, actionality, and Aktionsart. Wikipedia equates two of these and treats them as linguistic by saying that “The lexical aspect or aktionsart of a verb is a part of the way in which that verb is structured in relation to time” (, retrieved December 19, 2010). But it is possible that there are in fact two different, if interactive, systems of categories, one linguistic, the other extra-linguistic.

To a certain extent, lexical aspect constrains grammatical aspect; for example, English cannot normally use the progressive with a stative predicate (*John is being 4 years old; *A is following B in the alphabet). At the same time, aspect may serve to transform (“coerce”) one Aktionsart into another (cf. de Swart, this volume, §4.4): she climbed the mountain is eventive (she climbed the mountain in three days) but she was climbing is not (*she was climbing the mountain in three days).

Central issues in the study of (verbal) aspect include these:

  1. 1. What is grammatical aspect? How does it differ from lexical aspect?

  2. 2. How are the aspects and Aktionsarten defined, and how do they relate to one another?

  3. 3. How does grammatical aspect interact with lexical aspect on the one hand and tense on the other?

  4. 4. Is grammatical aspect universal, and why would we expect it (not) to be?

Question #1 is purely contingent and derives from the rather confused (and confusing) history of the study of “aspect.”8 Questions #2 and #3 are descriptive in nature, though they raise additional theoretical issues. For example, it has generally been assumed, and seems plausible, even likely, that in some sense the continuous and habitual meanings often marked by the same form are aspects of a broader semantic category of imperfectivity, but of course the question is what, if anything, in a comprehensive and adequate theory of aspect would predict this. The theoretical basis for the observation that tenseless languages make up for their lack of tense with aspect (see the chapter by Lin), and for the apparent universality of aspect is likewise questioned in #4. Finally, aspect is not only significant on the clausal and (p. 33) sentential levels, but enters into the study of text and discourse (as shown in the chapters by Fludernik—especially in §2, and Carruthers—especially in §§2.2, 4). The question then (#5) is how aspect on the intrasentential level relates to aspect on the extrasentential level. Answering questions such as these is the goal of the chapters in this part.

Lexical Aspect (Hana Filip)

Filip writes (§1) that lexical aspect is a semantic category that concerns properties of eventualities (in the sense of Bach, 1981) expressed by verbs. In the most general terms, the properties in question have to do with the presence of some end, limit or boundary in the lexical structure of certain classes of verbs and its lack in others. She is in accord with what de Swart says in her chapter (§1): “the thematic relation between the verb and its arguments is relevant to the aspectual characterization of the sentence”—read represents quite a different kind of thing than does read a story: *She read in five minutes, she read a story in five minutes; she read for five minutes, ?she read a story for five minutes; and ?she finished reading, she finished reading a story.

As we have seen, the existence of classes of eventualities differing in their ontological properties and of the corresponding categories of linguistic expressions differing in semantic properties raises a number of questions concerning the nature of both the linguistic and extralinguistic entities, as well as the bases and universality of the two classifications.

The various types of eventualities differ in their phasic properties and have sometimes been defined in phasic terms. States such as existing consist of a single, uniform phase. Instantaneous events such as spotting a coin on the pavement, Vendler’s “achievements,” are momentary transitions from state to state, utterly lacking duration. Other types of occurrences consist of distinct phases: Vendler’s “accomplishments” (like climbing a mountain) combine an “activity” phase (climbing on a mountain) with a point of culmination which is an instantaneous event (reaching the top).

Eventualities differ too in their mereology. States are like masses, cumulative (if you were ill from 3 to 5 and from 5 to 7, you were ill from 3 to 7) and divisible (if you were ill from 3 to 7, you were ill during any interval, and at any point in time, within that interval, hence the “subinterval property”9 of stative expressions). Event predicates (this term corresponding here to “telic predicates,” i.e., it is used as in Mourelatos, 1978/81 and Bach, 1981, 1986, for instance) are like count nouns in denoting individuals that lack cumulativity and divisibility, and allowing adverbs of qualification to be used with them as direct quantifiers over individuated events in their denotation (she admonishes her children a lot can mean “often”, but she appreciates her children a lot can only mean “to a great extent”, not “often”).

Another way of looking at eventualities is in terms of a number of—in most theories, three—parametric features: telicity, the possession of an inherent boundary; duration or extent; and dynamicity or change—or their lack. Although three binary (p. 34) features should define eight possibilities, most theorists have recognized only between three and six classes, such as the four in Filip’s list in her (9), which summarizes Bach’s (1986) classification: state, process (“activity”), protracted event (“accomplishment”), and momentaneous event (“achievement”).

The interaction of lexical and grammatical aspect has traditionally been characterized as involving constraints on co-occurrence: e.g., non-dynamic eventualities do not occur in the progressive (*the children are being young). In recent times, the relationship has rather been seen as one of type coercion, in which one type of expression is implicitly taken as denoting a different type of expression (thus, the progressive in he is dying “turns” an achievement into an accomplishment). How and why type coercion occurs has been the subject of much as yet inconclusive research.

Verbal Aspect (Henriëtte de Swart)

The distinction between tense and aspect on the one hand and between grammatical aspect and lexical aspect on the other, comes up in several chapters of this book. De Swart devotes section 1 to these questions.

Tense is essentially a deictic category, having to do with the relationship between times, traditionally the time of an eventuality E (an event, action, or state), and the time S of the speech act, functioning as a time of orientation or evaluation as well as a time of utterance or enunciation. This is what de Swart means, with reference to Hewson’s chapter, by verbal tense serving “to anchor the situation described by the sentence to the time axis.”

Verbal or grammatical aspect is also called “viewpoint” aspect, in contrast to lexical aspect, which is also called “situation” aspect (terms which come from Smith, 1991). The very same situation in the real world may be reported using different aspects (cf. (c), (d) below). Unlike tenses, which denote objective and mutually incompatible distinctions ((a) and (b) below are contradictory), aspects, while somehow meaning different things, do not mark objective distinctions, and there is no contradiction between (c) and (d). The difference is in how the situation in question is viewed—hence the Russian term vid “view” and its English gloss, aspect.

  1. (a) John won the race for the first time.

  2. (b) John will win the race for the first time.

  3. (c) I was sleeping all night.

  4. (d) I slept all night.

While verbal aspect is grammatical and part of the system of a particular language, lexical aspect “bears on inherent features of the verb,” as de Swart says (§1) and hence is universal by definition.

At the same time, while verbal aspect may be, and classically is, defined as concerning how the eventuality is viewed (or presented), so that in many cases more than one aspect can be used, for example the situation represented by the progressive in (c) and the perfective or neutral aspect in (d), in many, if not most, cases the (p. 35) choice of aspect is not facultative, either because there is an objective difference in meaning or the context forces a choice on the speaker (cf. (e, f), (g, h)).

  1. (e) When Suzanne came in, the baby was screaming.

  2. (f) When Suzanne came in, the baby screamed.

  3. (g) Mr. Blanding builds his dream house.

  4. (h) Mr. Blanding is building his dream house.

However we may wish to characterize aspect in general or define the various aspects, one thing that is true of grammatical aspect that is not true of either tense or lexical aspect is its compositionality. A clause may receive at most but one tense. The effect of piling morphosyntactic tense markers atop one another is not to refine the tense of the clause, but to alter it. Sue left, Sue has left, and Sue will have left represent three different tenses. But if Sue is leaving and Sue has been leaving represent different aspects (the progressive and the perfect progressive), it is in a different sense, since (i) and (j) are non-synonymous and potentially contradictory, while (k) and (l) are not, and may freely be used to describe one and the same situation.

  1. (i) Sue left today.

  2. (j) Sue will have left today.

  3. (k) Sue has slept all day.

  4. (l) Sue has been sleeping all day.

The relationship of tense, aspect, and Aktionsart raises the question of how the three interact on the syntactic and semantic levels. As de Swart notes in section 4.1, “Many current theories adopt some version of a layered representation in which tense syntactically and/or semantically dominates grammatical aspect, which in turn dominates aspectual class,” as diagramed in her example (32) (slightly modified here). This indicates that tense has greater scope than aspect, and aspect than Aktionsart (aspectual class).


[Tense [Aspect [aspectual class]]]

This is certainly not necessarily reflected in the surface structure of clauses; in (k, l), for example, the aspectual root have occurs before the tense ending. But in many syntactic theories, at some deep or underlying level of structure, possibly that of logical form, tense (T) c-commands a structure headed by aspect (Asp), which in turn c-commands the verb phrase proper (Figure 1).

As de Swart points out, there is considerable controversy over (32). First, is T universal, with so-called “tenseless” languages simply having a null tense? (Cf. the discussion in Lin, this volume, §3.) For that matter, is the constituent labeled Asp in Figure 1 in universal? It may be an empirical fact that it is found in all languages, but even if it were not, would a particular theory of syntax force a null aspect on one? Beyond that, as de Swart points out (§4.1), there are quite different configurations possible within the VP, “giving rise to a range of possible functional structures.”

(p. 36) Yet another question considered by de Swart is the semantic relationship between aspect and Aktionsart. One prominent view is that aspect functions precisely to transform the Aktionsart via aspectual coercion (§4.4).

De Swart’s chapter contains sections on the perfective/imperfective distinction in Slavic (§3.1) and French (§3.2) and on the progressive and perfect of English (§3.3), forming a preface to the chapters by Govozdanović and Mair which follow.

Perfective and Imperfective Aspect (Jadranka Gvozdanović)

The perfective/imperfective distinction looms large in aspectology. It was in connection with this distinction in Slavic that the concept of “aspect” first arose, and then came to the attention of Western scholars. It is also probably the most common aspectual distinction, grammaticalized in over 40% of the world’s languages (Mair, this volume). Finally, in Comrie’s seminal typology of aspects (1976, p. 25), this opposition—the distinction of a global, “external” view of a situation and a partial, “internal” view—is represented as the most fundamental one.

There are several problems associated with this characterization of the opposition, however. In Slavic, the imperfective is the unmarked member of the opposition (cf. Andrews, this volume). In English, however, it is the progressive—for Comrie, a sub-sub-type of the imperfective—which is marked. What this says about the universality of either aspect or markedness, or the relationship between markedness and meaning, remains contentious.

The relationship of perfectivity, a property of the predication, to telicity, a property of the predicate, poses another problem. The Slavic prefixes which alter aspect in most cases also change the meaning of their stems. For example, Russian imperfective pisat’ means “to write” but the perfectivized perepisat’ is “to copy”. Aspectual suffixes do not change meaning: perepisyvat’ is the imperfective of “to copy”. The (p. 37) question then is how telicity and perfectivity relate and whether the Slavic prefixal aspect is inflectional or derivational in nature.

Gvozdanović follows Gehrke (2007) in distinguishing between internal affixes, which may affect the argument structure controlled by the verb and induce telicity (inherent end-boundedness), and external affixes which affect the global situation and in principle do nothing but change the aspect. Internal affixes have only the verb within their scope, but external ones the entire verbal complex. Gvozdanović states, however, that affixes are not inherently internal or external, that this distinction is a result of the combination of affixal meaning with that of the verb base.

Against this background, Gvozdanović observes that there are two approaches to (im)perfectivity in Slavic aspectology: (1) the aspectual distinction concerns the (non)attainment of the inherent boundary of situations (being either the telic boundary or a temporal boundary), or (2) it concerns temporal constituency, whether the time over which the situation in question obtains is contained within (imperfective), or contains (perfective), the temporal frame (Reichenbach’s (1947) reference time, Klein’s (1994) topic time). Both views have their places outside Slavic studies as well.

Gvozdanović reports systematic differences in the use of aspects between eastern and western groups of Slavic languages, the latter using the perfective aspect more frequently than the eastern (Dickey, 2000). This usage difference, attributed by Dickey to a different conceptualization of the perfective in the two groups is illustrated in Gvozdanović’s Figure 2: western Slavic bases it on the autonomous event in its totality, the eastern, on its “temporal specificity which emerges in sequential relations.” However, in her analysis of parallel texts, she discovers that this fails to adequately account for Russian usage, and points (§1.4) to interactions of tense and aspect as also relevant.

She rejects as inadequate both Klein’s treatment based purely on temporal inclusion relations between situation time and topic time, and Dickey’s concept of temporal specificity, and develops a more refined account, which she then tests (in §2) through the scope of temporal quantifiers and verb prefixation in Russian and Czech.

Progressive and Continuous Aspect (Christian Mair)

As we have seen, in Comrie’s categorization of the aspects, the imperfective has two varieties, the continuous aspect and the habitual aspect, represented respectively by the two English glosses of the French imperfect tense: Jean lisait is either “Jean was reading” or “Jean used to read, Jean would read.” For Comrie, the continuous in turn has two varieties, the progressive (in which there is change over time) and the non-progressive, sometimes called the “continuous.” As Mair notes (§§1, 2), the demarcation between the two is problematical in practice, as is the distinction between the continuous (e.g., the horses had their hooves padded because they were kicking) and the habitual (the horses had their hooves padded because they kicked).

(p. 38) Nor is there a complete correspondence between the expression in a particular language of the concept of dynamic progression, and use of the progressive aspect—progression may be expressed in some other way, and progressive aspectual markers may have non-progressive uses, as in Mair’s example (8), I can only add that when Paul Gascoigne says he will not be happy until he stops playing football, he is talking rot.

Progression is expressed by marking the verb (as in Turkish Evimde çalışıyor “(s)he is working in my house,” example (1a)), by the use of periphrastic constructions, as in German Er ist einen Brief am Schreiben “he is writing a letter” (3c), or through particles, as in Mandarin Chàhn Sāang taan-gán saigaai “Mr. Chan is having a good time” (5). The typology of progressive constructions is the topic of section 2, in which Mair concludes that the morphosyntactic encodings of progressive and continuous aspect in individual languages vary widely in obligatoriness and grammaticalization.

Languages may be what Mair calls “abstract, de-contextualized structural systems,” but as noted in Fludernik’s chapter, they nonetheless function within, and in part are shaped by, human discourse. The questions Mair discusses in section 3, illuminated by recent changes in the English progressive, overlap with the concerns of both Fludernik and Carruthers in their chapters: What are the discourse implicatures of the progressive? If the choice of aspects is not determined semantically, what stylistic and sociolinguistic factors play a role in the choice? Is textual genre or medium an important factor in the selection? (This section also relates directly to the concerns of Nicolle’s chapter.)

Mair concludes (§4) that the degree of grammaticalization of progressives is highly variable, cross-linguistically, and that languages such as English, in which the category is largely obligatory, are in the minority. Interesting in the context of the notion of “actualization” (Nicolle, §2), the drastic increase in the discourse frequency of progressives across most genres that Mair finds in recent English is located mainly in the long-established uses, and is not due for the most part to the addition of new forms or functions. He attributes the change to shifting preferences at the discourse level and other pragmatic and extra-linguistic factors, but withholds cross-linguistic generalizations pending further research.

Habitual and Generic Aspect (Greg Carlson); Habituality, Pluractionality, and Imperfectivity (Pier Marco Bertinetto and Alessandro Lenci)

A second variety of the imperfective, according to Comrie (1976) and a plurality, if not a majority, of scholars since, is habitual aspect. Insofar as it is seemingly defined by repetition, it resembles, and sometimes has been confused with, the iterative Aktionsart, characterized by what Bertinetto and Lenci call (§1) “event-internal pluractionality,” e.g., Yesterday at 5 o’ clock John knocked insistently at the door. This must be contrasted with “event-external pluractionality,” which refers to the repetition of the same event in a number of different situations (e.g., John swam daily in (p. 39) the lake). Carlson (§1) notes that semelfactives10 such as knock and flap (cf. stroke below) readily receive an iterative interpretation. So-called “frequentatives”11 in languages such as English (e.g., flitter, waggle), Latin (cantare “to sing”, cursare “to run around”),12 and Russian (видывaть, “to see repeatedly”; пoглaживaть, “to stroke”) are probably best described as iterative.

On the other hand, habitual aspect has much in common with generic aspect. Devices dedicated to marking genericity are rare (Carlson, §1), it is often expressed by the same means as habituality, and ambiguities allowing either interpretation are possible, cf. Carlson’s example (1), The lion roars. The habitual reading concerns a specific lion involved in specific acts of roaring, whereas the generic reading concerns lions in general without reference to specific occurrences.

These three categories—iterative Aktionsart, and the habitual and generic aspects—provide the subjects of the chapter by Carlson, and that of Bertinetto and Lenci.

These two chapters propose alternative views of habituality. Bertinetto and Lenci claim that habituality is closely connected with the imperfective aspect and belongs to the domain of “gnomic imperfectivity,” whereas Carlson suggests that habituality may not be part of aspect at all: the problem of habituality is still a matter of debate. Hopefully, these two contributions will contribute to fostering further investigation.

Section 1 of the Bertinetto/Lenci chapter summarizes the grammatical and lexical devices used to mark repetition of occurrences, and various meaning distinctions so marked, e.g., reduplicativity (Italian riandare ‘to go again’), frequentativity (Lithuanian aš būdavau ‘I used to go’, Klimas, 1984), capacitativity (this engine vibrates, said of an engine which is not running), etc.

In section 2, Bertinetto and Lenci also insist that habituality belongs, in their view, to the imperfective aspect. They thus distinguish habituality—which they describe as “present[ing] a situation … as a characterizing property of an individual … during a given interval,” as in (b, their 1b)—from “iterative” perfective sentences, merely “present[ing] a plain state of affairs,” “establish[ing] a relation between an individual … and a time-interval,” as in (a, their 1a). Thus although (c, their 4c), in the French passé composé tense, involves repetition, they term it iterative, not habitual, unlike (d = 4d), in the imparfait.13 Both habituals and iteratives are pluractional—in the sense of “event-external pluractionality”—but they crucially differ in terms of aspect viewpoint (hence, of semantic interpretation). They note that adverbials of habituality are perfectly compatible with habitual sentences such as (e = 5b), but much less … so iterative ones (f = 5a).

  1. a. In the past few years, Franck has often taken the 8 o’ clock train.

  2. b. When he lived in the countryside, Franck would usually take the 8 o’ clock train.

  3. c. Louis a écrit cinq lettres. “Louis wrote/has written five letters.”

  4. d. Louis écrivait cinq lettres. “Louis wrote/would write/used to write five letters.”

  5. e. D’habitude, Olivier écrivait des poèmes. [habitual]

    (p. 40) “Usually, Olivier wrote [imperfective] poems.”

  6. f. ??D’habitude, Olivier a écrit des poèmes. [iterative]

    “Usually, Olivier wrote [perfective] poems.”

They further note that iterativity (g = 7b) is incompatible with the present, whereas habituality (h = 7d) is not. They observe too that “the framing adverbials of iterative and habitual sentences do not share the same constraints.”

  1. g. * Luc perd son parapluie trois fois.

    “Luc loses his umbrella three times.”

  2. h. Chaque année, Luc perd son parapluie trois fois. [habitual]

    “Every year, Luc loses his umbrella three times.”

In section 3, they discuss what they call “gnomic imperfectives,” which they divide into these categories (i-m, modified from 15a-e):

  1. i. At that time, John would easily get angry with his colleagues. [habitual]

  2. j. John smokes cigars. [attitudinal]

  3. k. John speaks Swahili. [potential, capitative]

  4. l. Elina is Finnish. [individual level (IL) predicate]

  5. m. Dogs have four legs. [generic]

Bertinetto and Lenci argue that habituals (e.g., i) differ from generics (m) because the former, but not the latter, necessarily involve (event-external) pluractionality; habituals are, in their view, at the intersection of pluractionality and imperfective gnomicity. That is, habituals entail the repetition of occurrences (like perfective “iterative” sentences, but contrary to generics), but at the same time they characterize a specific individual or situation over a period of time (like generics, but contrary to perfective “iterative” sentences). Generics, like IL-predicates (l) are intrinsically stative. By contrast, they argue, attitudinals (j) and potentials (k) receive stative interpretations via coercion from actional meanings.

Carlson quotes Comrie (1976, p. 27; 1985, p. 39) in noting that the repetition of occurrences is neither sufficient nor necessary for the use of habitual aspect: a habitual situation can be a “characteristic situation that holds at all times” (the temple of Diana used to stand at Ephesus) and “the mere repetition of a situation is not sufficient for that situation to be referred to by a specifically habitual … form.”

The formalization of the semantics of habituals and related sentences is the subject of section 4 of Bertinetto and Lenci’s chapter. They contrast (§4.1) analyses involving a monadic gnomic operator—which they characterize as “a sort of covert, default quantificational adverb,” so that the logical form of John always smokes after dinner differs from that of John smokes after dinner only in the occurrence of Always instead of the operator GEN—with relational models employing a dyadic operator. The logical form of sentences in this model involves a restrictor and a matrix clause, the former specifying “the conditions under which the state of affairs expressed in the matrix-clause hold.” In this case, John smokes is analyzed as something like GEN(e) [normal_smoke_situation(john,e)][smoke(e,john)]. The interpretation of (p. 41) the operator, and specifically the opposition between extensional and intensional treatments, is the subject of section 4.2. Bertinetto and Lenci argue for an intensional interpretation.

Although the habitual is traditionally regarded as an aspect, and Carlson continues to use the term habitual aspect, he casts doubt (§2) on its aspectuality. Although widely marked by imperfective aspectual markers, “it is not apparent from a distributional point of view that habituals form a systematic part of aspect systems from language to language, nor a part of tense systems, nor a part of any other systems in the verbal complex.” In section 8, he discusses analyses treating habituality as an aspect distinct from gnomic imperfectivity in general.

In sections 3 and 4, Carlson discusses the question of just what constitutes a marker of habitual aspect and argues (§3) that will/would is the sole marker of habituality in English, though used to is “consistent” with habituality. Noting (§4) that “the majority of languages do not employ a general, systematic means of expressing habituality,” he reports a wide range of devices for doing so in various languages, though he reports “some very strong patterns”: “Imperfectives …, progressives, inceptives, statives and continuatives appear most readily able to also express habituality,” and “the regular appearance of specifically habitual past tenses (akin to English ‘used to’)” contrasts with the lack of “genuine generic future forms.”

In section 7, Carlson tackles the issue of the meaning of habitual expressions, the multiplicity of specific notions encompassed by the rubric “gnomic imperfectivity” and the putative ambiguity of sentences such as John drinks beer, citing analyses in which the “propensity” reading receives a semantic analysis requiring existential quantification over times or possible worlds, while the habitual reading requires universal quantification, but also other analyses in which the difference is not semantic. It is relevant in this context that even languages with “habitual” markers allow other markings of habituality or genericity, but that there generally are differences in meaning between them.

In the end (§8), he points to a lack of strong universal patterns and concludes that “work on habituality and how it and similar notions [relate] to gnomic imperfectivity is still under development, and the questions outnumber the answers by quite some margin.”

Perfect Tense and Aspect (Marie-Eve Ritz)

The English present perfect tense is a puzzle, since, unlike the other perfect tenses, it appears to be non-compositional (see Ritz, §1). But as Ritz points out, the perfect itself is problematical, since (1) its form/meaning relation is cross-linguistically highly variable, (2) its meaning(s) and use(s) are uncertain, (3) it is diachronically unstable and the processes of its change are not well understood, and (4) its analysis has attracted a great many distinct approaches. We must even question whether it is in fact an aspect.14

(p. 42) Classically (McCawley 1971, 1981; Comrie, 1976) the perfect has been taken to have four distinct uses: (i) the perfect of “persistent situation” (universal perfect), (ii) the experiential (existential) perfect, (iii) the stative or perfect of result, and (iv) the perfect of recent past (McCawley’s “hot news” perfect). There has been a lively debate on whether these constitute different meanings.

The semantics of the perfect has raised other controversies as well (§§2–3), including whether it is a tense or an aspect, and whether there is such a thing as a “perfect” state, resultant from the eventuality implicit in the use of the perfect. Ritz surveys various theories.

Section 4 discusses its pragmatics, contrasting two theories, those of Portner (2003) and Nishiyama and Koenig (2004), both of which attempt to connect the meaning of the present perfect with the discourse context in which it occurs.

Other studies, examining the use of the perfect in relation to rhetorical (or “discourse”) relations, such as narration or result, are also discussed in section 4. De Swart (2007) concludes that the perfect establishes an elaboration structure, where speech time/the utterance situation is the topic. She argues that in general, sentences in the present perfect are related by a relation of continuation, which is neutral with regard to temporal ordering. In English and Dutch, no temporal relation is possible between eventualities, thus precluding use of the perfect in narration.

Section 5 summarizes the status of the four problems Ritz started with. Diachronic analysis has begun to provide further understanding of the sources for the diversity of meanings expressed by perfects cross-linguistically. Focus on the uses of the perfect in discourse has provided evidence for a range of textual functions, thereby providing a better understanding of the pragmatic factors involved. As for semantics, recent proposals seem to be converging on a picture of the present perfect as referring to the consequences or post-phase/state of an event. The four types of perfect, Ritz says, can be explained on the basis of the lexical aspect of the VP, the use of adverbials, and contextual factors. Rather than the variety of approaches hindering research, she sees research as having benefited from it.

Nonetheless, as in the case of the progressive, little of a general and definitive nature can as yet be said about the perfect, especially diachronically. Ritz calls for further research, especially typological, as well as research into sociolinguistic and discourse factors affecting the perfect.

Resultative Constructions (John Beavers)

In recent Western European and North American linguistics, the term “resultative construction” principally refers to “a phrase that indicates the state of a[n object denoted by a] noun [phrase] resulting from the completion of the [action denoted by the] verb [phrase]” (, retrieved December 18, 2010) such as wipe the counter clean; be drained dry; swing shut; shout oneself hoarse. Otherwise the term is more familiar as a tag for constructions associated with the name of Nedjalkov (1983/1988), e.g., be torn (the cloth is torn), havedone (he has the job done), etc. This chapter concerns the former type.

(p. 43) What the two types of “resultatives” have in common is the expression by the complement phrase of a state resultant from a previous action, which is expressed by the secondary predicate in Beavers’s type, but is merely implicit in Nedjalkov’s (as in the perfect).

The interest in resultative constructions comes precisely from this “event composition” resulting in a derived eventuality of the sort normally found in lexical accomplishments, though analysts differ regarding the nature of the composition (§1). The questions raised by their analyses touch on many of the issues elsewhere in this volume, involving compositionality, grammaticalization, the structures of eventualities, etc.

Section 2 concerns the various types of resultatives; e.g., the result phrase may be an Adjective Phrase (hammered the metal flat), but not a participial (*hammered the metal flattened). The specific types of constructions are constrained by the types of predicates: the freeze solid type requires an unaccusative15 verb, unergatives16 take a dummy reflexive object (yell oneself hoarse).

There has been a considerable amount of literature on resultatives, generally focused on the cross-linguistic availability of resultatives, variation in the inventories of construction types among languages that have resultatives, and the relationship of resultatives to other expressions of complex events, such as serial verbs and V-V compounds.

The key issue (§6) is how the verb and the result phrase (XP) combine to express a single event. In section 3, Beavers lays out the “Classic Accomplishment Analysis.” A leading assumption has been that resultatives represent derived lexical accomplishments. Evidence for this comes from the fact that resultatives and lexical accomplishments pattern together in terms of telicity, adverb scope, and causation diagnostics. Likewise, the specific causal relation for both resultatives and lexical causatives must be “direct” causation, with no event intermediate between the causative event (e.g., hammering) and the result state (being flat).

In section 4, he discusses a number of problems for the classic view. Beavers points out that not all resultatives are causal in nature; in particular, resultatives with unaccusatives (e.g., freeze solid), resist causative paraphrases.

Resultatives that do have causal readings do not always involve causation of the XP-denoted state by the verb-denoted event (28), and causation may run the other way (29).


Smith cut the bread into thick slices. (Rapoport, 1999, p. 671, (42a))


The bullets whistled past the house. ≠ ‘The bullet’s whistling caused it to be past the house.’ (Cf. The bullet’s moving past the house caused it to whistle.)

Beavers discusses a novel approach to event composition, proposed by Rappaport Hovav and Levin (2001), in which the crucial relationship between the verb and the XP is defined in terms of temporal dependency between the events denoted by the V and the XP. While it solves some of the problems of the classic approach, it has its own problems; in particular, it cannot account for a causal relationship in those cases in which it occurs. He also discusses temporal dependency approaches (p. 44) proposed by Wechsler (2005) and Beavers (2008) in which the XP denotes a scale that homomorphically delimits the end of the change event. This approach makes many aspectual and grammatical predictions not captured by previous approaches, but fails to explain recently noted cases of resultatives that are atelic or that do not entail actual change. Beavers concludes that solutions in terms of temporal dependency still leave many questions unanswered, but suggests that future work on derived eventualities may want to take into account more recent advances on types of lexicalized eventualities.

Aspect and Diathesis

The next two chapters concern two areas often neglected in the study of aspect, but which clearly are relevant to it, first in terms of the marking of aspectual distinctions, but also in terms of the interaction of aspectual and temporal semantic categories with diathetic ones.

Voice (Mila Vulchanova)

The term diathesis, sometimes used as a synonym for grammatical voice and sometimes as its superordinate term, is unfamiliar even to many grammarians. Vulchanova accordingly encompasses the sundry diathetic alternations within a broad use of the term “voice.”

Vulchanova’s chapter principally concerns the relationship to voice of aspect (§§1–3) (but also of tense, in §§4, 5). Morphologically, the two may be marked in related ways, as in the case of perfect/passive participles such as broken. Semantically, both voice and aspect relate—as morphological case largely does (cf. Richardson, this volume)—to the structure of a situation and to information regarding the participant roles associated with the verb (§1).

Voice may be viewed, Vulchanova says (§2), as the range of potential alternative realizations in syntax of the same verbal root. Aspect construal depends not only on the mapping of both internal and external arguments, but also on their quantificational properties, as in Verkuyl’s influential theory (cf. his chapter in this volume). She discusses grammatical processes and constructions closely related to aspect construal which affect diathesis, e.g., resultatives (cf. Beavers, this volume) and causatives, asking (a) how these are generated, (b) what allows for the generation in the first place, and (c) what constraints there are on the generation. In this section she describes two major types of approaches that have emerged: constructionist and lexicalist.

She reports that most recent work has addressed aspectual values in the context of what are considered as canonical or default realizations of arguments such as active sentences in Indo-European (§2). Particularly in ergative languages, aspect has been shown to strongly correlate and interact with the mapping of arguments to syntax.

(p. 45) Section 3.1 concerns whether a diathetic alternation yields a new aspectual value, and the relationship between resultative constructions, the passive, and the perfect. Section 3.2 concerns the extent to which aspect construal is constrained, e.g., by passive morphology. Vulchanova concludes that “one can expect that various diathetic forms differ as regards the possibility of aspect construal, and are further constrained by the lexical semantics of the head verb and more specifically, by the kind of process it denotes and what kind of (internal) argument this process applies to.”

In §4, Vulchanova explores cases in which tense places restrictions on diathetic alternations, such as passives. For example, the Germanic middle tends to occur in the present tense and expresses a generic value.

In §5, she discusses the question of what gives rise to the constraints on aspect construal and the expression of certain temporal categories in their interaction with diathetic alternations. Essentially, there are two approaches, one that attributes these restrictions to originally lexical properties of the respective components, and another that attributes this to coercion by the specific syntactic structures in which these components are embedded. Vulchanova opts for an approach that recognizes the participation of both, at different levels of the composition.

As in other areas, no definitive answers have been found for the central questions raised in this chapter, reflecting the complexities not only of aspect, but also of its interactions with voice on the one hand, and temporal categories on the other (§6). It does not help that the field has suffered from a “[pronounced] proliferation of ideas, frameworks, formats, and terminology,” aggravated by the differences between, for example, the Slavicist and other traditions of study (§1.2).

Case (Kylie Richardson)

The relevance of case to aspect goes beyond the well-known use of cases to mark aspectual relationships in Finnish, which Richardson (§4) describes as “exhibit[ing] an accusative-versus-partitive-case opposition on an internal argument that appears to be linked to an aspectual contrast” as illustrated by examples (4) and (5):





“I shot at the/a bear.”





“I shot the/a bear.”

In this chapter, Richardson aims (§6) to show that “morphological case [discussed in her section (§2)] is aspectually relevant across a wide range of different languages,” notwithstanding the difficulty of establishing a one-to-one correlation between case and aspectual features (§1). The languages cited include, among others, the Slavic languages, two Germanic languages (German and Icelandic), some Uralic languages (Finnish, Estonian, Hungarian, Inari Saami), Indic languages (Hindi, Bengali), an Australian language (Walpiri), and a Sepik language of Papua New Guinea (Manambu).

(p. 46) She points out (§1) that morphological case has been connected to the “lexical/semantic” aspectual feature telicity and/or to the “grammatical/morphological” one of boundedness. The two types of aspect—lexical or situation aspect and grammatical or viewpoint aspect—are discussed in §3, specifically in connection with the relationship of boundedness to telicity, and Richardson argues that “the defining characteristic of an imperfective or perfective eventuality as unbounded or bounded in time is distinct from the notion of telicity,” though case is relevant to both boundedness and telicity, and hence to both kinds of “aspect.” (The issue of the relationship of perfectivity to telicity is also discussed in the chapters by Deo and Gvozdanović.)

For example, she reports (§4) claims that Finnish case is not only relevant to lexical/semantic aspect but to grammatical aspect: that the inessive and addessive (ADES) cases are linked to the imperfective aspect, while the elative, illative, ablative, and allative (ALL) cases are linked to the perfective, contrasting the examples (6, 7):


Juna pysähtyi


train stop-PAST-3sg


“The train stopped at the station (i.e., toward the station as its final destination).”


Juna pysähtyi


train stop-PAST-3sg


“The train stopped at the station (i.e., at the station as a passing point).”

She says that “[a]ccording to Sands (2000, p. 277), in example (6), focus is on the process of the eventuality described by the verb phrase, not the result, whereas in example (7), focus is on the result, e.g., this station may be the train’s final destination,” and adds that “the different interpretations of these two examples are reminiscent of the unbounded-bounded readings of the imperfective versus the perfective aspect in the Slavic languages.”

In section 4, she discusses the link between nominal case and grammatical aspect in the history of German and in a number of ergative-absolutive languages (e.g., Warlpiri and Hindi); the link between case on adverbial phrases and grammatical aspect in Inari Saami, Finnish, and Russian; the connections between case and lexical aspect in Latin, Classical Greek, Hebrew, Icelandic, and Manambu. In section 5, she discusses in detail the relationship of case-marking in Slavic to the semantics of verbal prefixation and the event structures of verb phrases.

However, the extent to which aspectual relevance can or should be captured syntactically, semantically, or otherwise, she leaves open (§6).


The third member of the TAM triple is sometimes identified as mood, sometimes as modality. Johnson (1981) speaks of a category she calls status, and defines tense, aspect and status alike in terms of the three points in time, S, E, and R. We (p. 47) have seen that tense can be viewed as the relationship between R and S, and aspect as that between E and R. This leaves the third member of Johnson’s triple, status, as relating E to S. Johnson is concerned with the status distinction of “Manifest/Immanent” (eventualities which are part of history vs. those which are not) in the Bantu language Kikuyu. Such a distinction is clearly related to both mood and modality.

Mood is a sub-system of grammar, but one difficult to define. The American Heritage Dictionary offers a fairly standard definition:

A set of verb forms or inflections used to indicate the speaker’s attitude toward the factuality or likelihood of the action or condition expressed. In English, the indicative mood is used to make factual statements, the subjunctive mood to indicate doubt or unlikelihood, and the imperative mood to express a command.17

Two problems are evident in this definition: first, mood is expressed in languages that lack verbal inflections, and second, “expressing a command” (that is, committing a speech act) is clearly a different sort of thing than “indicating the speaker’s attitude.”

Modality is a concept from semantics. The American Heritage Dictionary again offers a fairly standard definition: “The classification of propositions on the basis of whether they assert or deny the possibility, impossibility, contingency, or necessity of their content.”

Although mood and modality are closely enough related that mood is sometimes characterized roughly as the grammatical expression of modality (“Mood is one of a set of distinctive forms that are used to signal modality”),18 the two are independent, with mood being used at times to express other things than modality (for example, commands in the case of the imperative and wishes in the case of the optative) and modality being expressed in other ways than mood (for example, through the use of modal auxiliary verbs like can, could, may, might, etc.).

Tense and aspect relate closely to mood, as well as to modality, in terms of both their meanings and their markers. For this reason, while a fuller consideration is beyond the scope of this volume, something needs to be said about aspects of mood and modality that touch on tense and aspect, and two chapters here deal with some such aspects.

Time in Sentences with Modal Verbs (Ilse Depraetere)

It is useful to refer to TAM (tense-aspect-mood/modality) systems rather than to tense-aspect systems alone. While not directly time-related, modality on the semantic level and mood on the morphosyntactic level interact closely with both tense and aspect.

It is well known that tense markers may receive modal interpretations or trigger modal inferences, and that the converse is true as well. In this chapter, Depraetere is concerned specifically with the communication of temporal information in utterances with modal verbs, using clauses with epistemic and root possibility (hence with the verbs can, could, may, might) as her “test cases.”

(p. 48) She distinguishes (§1.2) two independent times in modal clauses: the time at which the modality holds (the “M-situation” and the time of the “residue,” the predicated situation which remains after the modality is stripped off. Thus she contrasts present (5), past (6), and future (7) possibility, and past (8), present (9), and future (10) “residues”:


She can swim.


She could swim at the age of five.


She will be able to swim after she’s taken swimming lessons.


You may have been right.


You may be right.


We may be back by ten.

The principal questions she seeks to answer are (§1.4):

  1. 1. What is the temporal location of the M-situation? (discussed in §2)

  2. 2. What are the temporal relationship between the M-situation and the residue in the cases of wide scope modality (e.g., in the case of epistemic modality) (§§2.1, 2.4) and when the modal meaning has narrow scope (e.g., in the case of ability and permission)? (§§2.2, 2.3)

  3. 3. What are other linguistic factors that further influence the temporal interpretation of utterances containing modals? (§3)

The defective morphology and diachronic development as preterite-present verbs have left the English modals with unusual semantics that plays an important role in their use and interpretation, as Depraetere notes (§1.5). Except in indirect speech constructions past tense modals may (18) or may not (19) have past time reference. In some cases, the use of a periphrastic form may be allowed (can = is able, is permitted) or is necessary (*will can: will be able).


At that time he might also by chance have met his sister Isabella


The primary object of a new police might be couched in the following or similar terms.

She concludes (§2.4.1) that the major factors determining the possible range of temporal relations between the M-situation and the residue are the scope and the nature of the modality.

The interactions of modality and tense are complex and a general theory of time expression in modal utterances will require much more investigation.

Evidentiality and Mirativity (Ferdinand de Haan)

Evidentiality is often, but not invariably, considered to be the expression of epistemic modality (cf. Palmer, 1986, pp. 51–54), though the connection has been questioned by, inter alia, de Haan (1999, 2001). Evidentiality—and the closely related mirativity (§7)—are also formally and semantically connected with tense and aspect. Some markers (such as the Turkish past tenses in (i)) are portmanteaus, (p. 49) expressing both tense/aspect and evidentiality; evidential markers (as in (iib) below) sometimes develop from tense or aspect morphemes (here the same marker as in the related Carapana language (iia)); and tenses and/or aspects may have evidential implications and vice-versa (iii, iv), e.g., Swedish has a present perfect (iii) considered to have inferential meaning (Thorell, 1973, §407; Kinnander, 1974, p. 129; cf. Haugen, 1972) while the mirative generally implies a recent or on-going occurrence (iv).


geldi “he/she/it came”, gelmiş “he/she/it apparently/reportedly came” (Slobin and Aksu, 1982, pp. 186–187)


(a) =

(de Haan’s 15)

Carapana (E. Tucanoan)


‘She worked.’ (no apparent evidential reading)


(de Haan’s 16)

Tucano (E. Tucanoan)


‘She was.’ (witnessed past)


På försommaren 1814 har Stagnelius säkerligen återvändt till hemmet i Kalmar.

“Stagnelius [most likely] returned to his home in Kalmar in the early summer of 1814.”


= (de Haan’s 38b)

Hare (Athabaskan)

heee, gúhde daweda! ch’ifi dach’ida lõ

“Hey, he is sitting up there! The guy is sitting up there.”

De Haan’s chapter explores the various ways in which evidentiality manifests itself in language, with discussions of the various types of evidentiality (the semantic side of evidentiality, in §3) and the ways evidential notions are expressed cross-linguistically (morphosyntactic expressions of modality, §4). The focus is on work in the functional-typological tradition, but other viewpoints receive some attention as well.

Evidentiality was formerly seen as an exotic feature of mainly non-Western, non-Indo-European languages found in the Americas and Asia (this despite Haarmann, 1970, on indirect evidentiality in the languages of Europe). However, in recent years, studies of evidentiality have shown that evidentiality does play a role in the languages of Europe, even though it may not be expressed in quite the same way as in, say, languages of the Americas such as Tuyuca. The history of the study of evidentiality is briefly reviewed in de Haan’s section 2.

Section 3 categorizes evidential/inferential markers, distinguishing direct (firsthand or sensory) evidentials that indicate that the speaker actually witnessed the occurrence and indirect (secondhand) evidentials, and explores relationships between these categories or their sub-categories. The direct evidentials include visual, auditory, and nonvisual evidentials. The indirect are quotatives (p. 50) (hearsay) and inferentials, which may be distinguished (as in Ainu) or not (as in Turkish).

Section 4 offers examples of various lexical and grammatical ways of marking evidentiality, including tense/aspect markers (§4.4) and mood (§4.5).


Abusch, D. (1993). Sequence of tense revisited: Two semantic accounts of tense in intensional contexts. In H. Kamp (ed.), Ellipsis, tense and questions (pp. 85–140). DYANA-2, Dynamic Interpretation of Natural Language, esprit Basic Research Project 6852, deliverable R2.2.B. Shorter version appeared in P. Dekker and M. Stokhof (eds.), Proceedings of the 9th Amsterdam Colloquium, 1993. Amsterdam: Department of Philosophy, University of Amsterdam. (Available at, retrieved February 22, 2011.)Find this resource:

Abusch, D. (1997). Sequence of tense and temporal de re. Linguistics and Philosophy, 20, 1–50.Find this resource:

Andersen, R. (ed.) (1983). Pidginization and creolization as language acquisition. Rowley: Newbury House.Find this resource:

Andrews, E. (1990). Markedness theory: The union of assymetry and semiosis in language. Durham, N.C.: Duke University Press.Find this resource:

Arends, J. (1989). Syntactic developments in Sranan: Creolization as a gradual process. Ph.D. dissertation, Catholic University, Nijmegen. (Available at, retrieved October 2, 2010.)Find this resource:

(p. 52) Arends, J. (1996). The early stages of creolization. Creole Language Library, 13. Amsterdam: Benjamins.Find this resource:

Bach, E. (1981). On time, tense, and aspect: An essay in English metaphysics. In P. Cole (ed.), Radical pragmatics (pp. 63–81). New York: Academic Press.Find this resource:

Batistella, E. L. (1990). Markedness: The evaluative superstructure of language. Albany: SUNY Press.Find this resource:

Bennett, M., & Partee, B. (1972). Toward the logic of tense and aspect in English. Santa Monica, CA: System Development Foundation. Reprinted, 1978, Bloomington: University of Indiana Linguistics Club. Reprinted, 2004, in Compositionality in formal semantics: Selected papers by Barbara H. Partee (pp. 59–109). Oxford: Blackwell.Find this resource:

Benveniste, E. (1959). Les relations de temps dans le verbe françis. Bulletin de la Societé Linguistique de Paris, 54, 69–82. Reprinted, 1966, in E. Benveniste, Problèmes de linguistique générale, I. Paris: Gallimard, Paris.Find this resource:

Bickerton, D. (1976). Creole tense-aspect systems and Universal Grammar. Cambridge: Cambridge University Press.Find this resource:

Bickerton, D. (1981). Roots of language. Ann Arbor: Karoma.Find this resource:

Bickerton, D. (1983). Creole languages. Scientific American, 249(8), 116–122.Find this resource:

Bickerton, D. (1984). The language bioprogram hypothesis. The Behavioral and Brain Sciences, 7, 173–221.Find this resource:

Bickerton, D. (1988), Creole languages and the bioprogram, in F. J. Newmeyer (ed.), Linguistics: The Cambridge survey, 2. Cambridge: Cambridge University Press.Find this resource:

Binnick, R. I. (1991). Time and the verb: A guide to tense and aspect. New York: Oxford University Press.Find this resource:

Bull, W. E. (1960). Time, tense and the verb. Berkeley: University of California Press.Find this resource:

Carlson, L. (2009). Tense, mood, aspect, diathesis: Their logic and typology. (Available at, retrieved September 7, 2010, and, retrieved February 22, 2011.)

Carnap, R. (1947). Meaning and necessity. Chicago: University of Chicago Press.Find this resource:

Carruthers, J. (2005). Oral narration in Modern French: A linguistic analysis of temporal patterns. Oxford: Legenda, monographs in French.Find this resource:

Chomsky, N. (1957). Syntactic structures. Janua linguarum, 4. The Hague: Mouton.Find this resource:

Chomsky, N. (2001). Derivation by phase. In M. Kenstovicz (ed.), Ken Hale: A life in language (pp. 1–54). Cambridge, MA: MIT Press.Find this resource:

Chomsky, N. (2008). On phases. In R. Freidin, C. Otero, & M.-L. Zubizarreta (eds.), Foundational issues in linguistic theory: Essays in honor of Jean-Roger Vergnaud. Cambridge: MIT Press.Find this resource:

Comrie, B. (1976). Aspect. Cambridge: Cambridge University Press.Find this resource:

Comrie, R. (1985). Tense. Cambridge: Cambridge University Press.Find this resource:

DeCaen, V. (1996). Tenseless languages in light of an aspectual parameter for universal grammar: A preliminary cross-linguistic survey. Toronto Working Papers in Linguistics, 14(2), 41–81.Find this resource:

de Haan, F. (1999). Evidentiality and epistemic modality: Setting boundaries. Southwest Journal of Linguistics, 18, 83–101. (Available at, retrieved February 22, 2011.)Find this resource:

de Haan, F. (2001). The relation between modality and evidentiality. In R. Müller & M. Reis (eds.), Modalität und Modalverben im Deutschen. Linguistische Berichte, Sonderheft 9. Hamburg: H. Buske. (Available at, retrieved February 22, 2011.)Find this resource:

(p. 53) Dever, J. (2006). Compositionality. In E. Lepore and B.C. Smith (eds.), The Oxford handbook of the philosophy of language (pp. 633–666). Oxford: Clarendon Press. (Available at A longer version is available at, both retrieved February 22, 2011.)Find this resource:

DeGraff, M. (ed.). (1999). Language creation and language change: Creolization, diachrony and development. Cambridge, MA: MIT Press.Find this resource:

Dickey, S. (2000). Parameters of Slavic aspect: A cognitive approach. Stanford: CSLI.Find this resource:

Dinsmore, J. (1982). The semantic nature of Reichenbach’s tense system. Glossa, 16, 216–239.Find this resource:

Enç, M. (1981). Tense without scope: An analysis of nouns as indexicals. Ph.D. dissertation, University of Wisconsin, Madison.Find this resource:

Enç, M. (1986). Towards a referential analysis of temporal expressions. Linguistics and Philosophy, 9, 405–426.Find this resource:

Enç, M. (1987). Anchoring conditions for tense. Linguistic Inquiry, 18, 633–657.Find this resource:

Engel, D. M. (1990). Tense and text: A study of French past tenses. London and New York: Routledge.Find this resource:

Fleischman, S. (1990). Tense and narrativity: From medieval performance to modern fiction. Austin: University of Texas Press.Find this resource:

Fludernik, M. (1993). The fictions of language and the languages of fiction: The linguistic representation of speech and consciousness. London and New York: Routledge.Find this resource:

Gehrke, B. (2007). Ps in motion. Ph.D. thesis, Utrecht University (LOT dissertation series, 184).Find this resource:

Greenberg, J. H. (1963). Some universals of grammar with particular reference to the order of meaningful elements. In J. H. Greenberg (ed.), Universals of language (pp. 73–113). Cambridge, MA: MIT Press. (Available at, retrieved February 22, 2011.)Find this resource:

Haarmann, H. (1970). Die indirekte Erlebnisform als grammatische Kategorie: Eine Eurasische Isoglosse. Wiesbaden: Harrassowitz.Find this resource:

Hamburger, K. (1957). Die Logik der Dichtung. Stuttgart: Klett. 2nd ed., 1968.Find this resource:

Haspelmath, M. (2006). Against markedness (and what to replace it with). Journal of Linguistics, 42(1), 25–70. (Ms. version available at Retrieved September 9, 2010.)Find this resource:

Haugen, E. (1972). The inferential perfect in Scandinavian, a problem of contrastive linguistics. Canadian Journal of Linguistics, 17, 132–139.Find this resource:

Jahn, M. (2005). Narratology: A guide to the theory of narrative. Cologne: English Department, University of Cologne. (Available at, retrieved February 22, 2011.)Find this resource:

Jespersen, O. (1924). The philosophy of grammar. London: Allen & Unwin.Find this resource:

Jiang, Z., & Shao, C. (2006). Markedness in universal grammar and second language acquisition. US-China Education Review, 3(8), 77–79.Find this resource:

Johnson, M.R. (1981). A unified temporal theory of tense and aspect. In P. Tedeschi & A. Zaenen (eds.), Syntax and semantics, 14: Tense and aspect (pp. 145–175). New York: Academic Press.Find this resource:

Jokić, A., & Smith, Q. (eds.). (2003). Time, tense, and reference. Cambridge: MIT Press.Find this resource:

Joos, M. (1964). The English verb: Form and meanings. Madison: University of Wisconsin Press.Find this resource:

(p. 54) Kinnander, B. (1974). Perfektum i sekundär användning. Nysvenska studier, 53, 127–172.Find this resource:

Klimas, A. (1984). Some unsolved riddles of Lithuanian linguistics. Lituanas, 30(1). (Available at, retrieved September 10, 2010.)Find this resource:

Klein, W. (1994). Time in language. London: Routledge.Find this resource:

Kondrashova, N. (2005). Is Russian a split-SOT language? Presented at the 14th Annual Workshop on Formal Approaches to Slavic Linguistics, Princeton University, Princeton, NJ, May 6–8. (Abstract available at, retrieved December 5, 2010).

Kortmann, B. (1991). The triad “tense-aspect-Aktionsart”: Problems and possible solutions. In C. Vetters et al. (eds.), Perspectives on aspect and Aktionsart (pp. 9–27). Bruxelles: Ed. de l’Université de Bruxelles. (Belgian Journal of Linguistics, 6.) (Available at, retrieved February 23, 2011.)Find this resource:

Kubota, Y., Lee, J., Smirnova, A., & Tonhauser, J. (2009a). Cross-linguistic variation in temporal adjunct clauses. Cahier Chronos: Selected Proceedings of Chronos 8. Amsterdam and Atlanta: Rodopi. (Available at Smirnova/Papers/124579/Cross-linguistic_variation_in_temporal_adjunct_clauses_ to_appear_;, retrieved December 5, 2010.)Find this resource:

Kubota, Y., Lee, J., Smirnova, A., & Tonhauser, J. (2009b). On the cross-linguistic interpretation of embedded tenses. In A. Riester & T. Solstad (eds.), Proceedings of Sinn und Bedeutung, 13 (pp. 307–320). Working Papers of the SFB 732. Stuttgart: University of Stuttgart.Find this resource:

Kučera, H. (1980). Markedness in motion. In C. V. Chvany & R. D. Brecht (eds.), Morphosyntax in Slavic. Columbus: Slavica.Find this resource:

Lecarme, J. (1999). Nominal tense and tense theory. In F. Corblin, C. Dobrovie-Sorin, & J.-M. Marandin (eds.), Empirical issues in formal syntax and semantics 2 (pp. 333–354). The Hague: Thesus.Find this resource:

Lefebvre, C. (1986). Relexification in creole genesis revisited: The case of Haitian creole. In P. Muysken and N. Smith (eds.), Substrata versus universals in creole genesis (pp. 279–301). Amsterdam: Benjamins.Find this resource:

Lefebvre, C. (1993). The role of relexification and syntactic reanalysis in Haitian Creole: Methodological aspects of a research program. In S. Mufwene (ed.), Africanisms in Afro-American Language varieties (pp. 254–279). Athens: University of Georgia Press.Find this resource:

Lefebvre, C. (1998). Creole genesis and the acquisition of grammar. Cambridge: Cambridge University Press.Find this resource:

Lefebvre, C., White, L., & Jourdan, C. (eds.). (2006). L2 acquisition and creole genesis: Dialogues. Amsterdam: Benjamins.Find this resource:

Le Poidevin, R., & MacBeath, M. (eds.). (1993). The philosophy of time. Oxford: Oxford University Press.Find this resource:

Lyons, J. (1977). Semantics. Cambridge: Cambridge University Press.Find this resource:

Mani, I., Pustejovsky, J., & Gaizauskas, R. (eds.). (2005). The language of time: A reader. Oxford: Oxford University Press.Find this resource:

McCawley, J. D. (1971). Tense and time reference in English. In C. Fillmore and T. Langendoen (eds.), Studies in linguistic semantics (pp. 96–113). New York: Holt, Rinehart and Winston.Find this resource:

(p. 55) McCawley, J. D. (1981). Notes on the English present perfect. Australian Journal of Linguistics, 1, 81–90.Find this resource:

McGilvray, J. (1974). Tenses and beliefs. Unpublished manuscript, McGill University.Find this resource:

McTaggart, J. M. E. (1908). The unreality of time. Mind, 68, 457–474.Find this resource:

McWhorter, J. H. (1998). Identifying the creole prototype: Vindicating a typological class. Language, 74(4), 788–818.Find this resource:

McWhorter, J. H. (2005). Defining creole. New York: Oxford University Press.Find this resource:

Młynarczyk, A. K. (2004). Aspectual pairing in Polish. Utrecht: LOT. (Available at, retrieved February 23, 2011.)Find this resource:

Moens, M. (1987). Tense, aspect and temporal reference. Ph.D. thesis, Centre for Cognitive Science, University of Edinburgh. Scotland.Find this resource:

Mufwene, S. (2001). The ecology of language evolution. Cambridge: Cambridge University Press.Find this resource:

Nedjalkov, V.P. (ed.). (1983). Tipologija rezul’tativnyx konstrukcij. Leningrad: Nauka. Translated, 1988, as Typology of resultative constructions. Amsterdam: Benjamins.Find this resource:

Nishiyama, A., & Koenig, J.-P. (2004). What is a perfect state? WCCFL, 23 (pp. 595–606). Davis, CA: Cascadilla Press.Find this resource:

Ogihara, T. (1995). Double-access sentences and reference to states. Natural Language Semantics, 3, 177–210.Find this resource:

Ogihara, T. (1996). Tense, attitudes, and scope. Dordrecht: Kluwer Academic.Find this resource:

Pagin, P., & Westerståhl, D. (2008). Compositionality. Forthcoming in C. Maienborn, K. von Heusinger, & P. Portner (eds.), Semantics. An international handbook of natural language meaning. Berlin: Mouton. (Available at, retrieved February 22, 2011.)Find this resource:

Pagin, P., & Westerståhl, D. (2010a). Compositionality I: Definitions and variants. Philosophy Compass, 5, 250–264. (Available at, retrieved February 22, 2011.)Find this resource:

Pagin, P., & Westerståhl, D. (2010b). Compositionality II: Arguments and problems. Philosophy Compass, 5, 265–282. (Available at, retrieved February 22, 2011.)Find this resource:

Palmer, F. (1986). Mood and modality. Cambridge: Cambridge University Press.Find this resource:

Partee, B. (1973). Some structural analogies between tenses and pronouns in English. The Journal of Philosophy, 70, 601–609.Find this resource:

Pelletier, F. J. (2001). Did Frege believe Frege’s Principle? Journal of Logic, Language and Information, 10(1), 87–114. (Available at, retrieved February 22, 2011.)Find this resource:

Portner, P. (2003). The (temporal) semantics and (modal) pragmatics of the perfect. Linguistics and Philosophy, 26, 459–510.Find this resource:

Rapoport, T. R. (1999). Structure, aspect, and the predicate. Language, 75, 653–677.Find this resource:

Rappaport Hovav, M., & Levin, B. (2001). An event structure account of English resultatives. Language, 77, 766–797.Find this resource:

Reichenbach, H. (1947). Elements of symbolic logic. New York: Macmillan.Find this resource:

Sands, Kristina. (2000). Complement clauses and grammatical relations in Finnish. Ph.D. dissertation, Australian National University, Canberra.Find this resource:

Selbach, R., Cardoso, H. C., & van den Berg, M. (eds.). (2009). Gradual creolization: Studies celebrating Jacques Arends. (Creole Language Library, 34.) Amsterdam: Benjamins.Find this resource:

Siegel, J. (1999). Transfer constraints and substrate influence in Melanesian Pidgin. Journal of Pidgins and Creole Languages, 14, 1–44.Find this resource:

(p. 56) Siegel, J. (2000). Substrate influence in Hawai’i Creole English. Language in Society, 29, 197–236.Find this resource:

Siegel, J. (2008). Pidgins/creoles and second language acquisition. In J. V. Singler & S. Kouwenberg (eds.), The handbook of pidgin and creole studies (pp. 189–218). Oxford: Blackwell.Find this resource:

Singler, J. V. (ed.) (1990). Pidgin and creole tense-mood-aspect systems. (Creole Language Library, 6.) Amsterdam: John Benjamins.Find this resource:

Singler, J. V. (1993). African influence upon Afro-American language varieties: A consideration of sociohistorical factors. In S. Mufwene (ed.), Africanisms in Afro-American language varieties (pp. 235–253). Athens: University of Georgia Press.Find this resource:

Slobin, D. I., & Aksu, A. A. (1982). Tense, aspect and modality in the use of the Turkish evidential. In P. J. Hopper (ed.), Tense and aspect: Between semantics and pragmatics (pp. 185–200). Amsterdam and Philadelphia: John Benjamins.Find this resource:

Smith, C. S. (1991). The parameter of aspect. Dordrecht: Kluwer.Find this resource:

Smith, C. S. (2003). Modes of discourse: The local structure of texts. Cambridge: Cambridge University Press.Find this resource:

de Swart, H. (2007). A cross-linguistic discourse analysis of the perfect. Journal of Pragmatics, 39(12), 2273–2307.Find this resource:

Sybesma, R. (2007). Whether we tense-agree overtly or not. Linguistic Inquiry, 38, 580–587.Find this resource:

Szabo, Z. G. (2000). Problems of compositionality. New York: Garland.Find this resource:

Szabo, Z. G. (2007). Compositionality. Stanford Encyclopedia of Philosophy. ( Retrieved February 22, 2011.)

Thorell, O. (1973). Svensk grammatik. Stockholm: P. A. Norstedt & Söer.Find this resource:

Tonhauser, J. (2006). The temporal semantics of noun phrases: Evidence from Guaraní. Ph.D. dissertation, Stanford University.Find this resource:

Trager, G. L., & Smith, H. L. Jr. (1951). An outline of English structure. Washington, DC: American Council of Learned Societies.Find this resource:

Vendler, Z. (1957). Verbs and times. The Philosophical Review, 66(2), 143–160.Find this resource:

Verkuyl, H. J. (1972). On the compositional nature of the aspects. Dordrecht: Reidel.Find this resource:

Verkuyl, H. J. (1989). Aspectual classes and aspectual composition. Linguistics and Philosophy, 12, 39–94. (Available at, retrieved February 22, 2011.)Find this resource:

Weinrich, H. (1964). Tempus: Besprochene und erzälte Welt. Reprinted, 1985 and 2001, Munich: Beck. Translated, 1973, as Le temps. Paris: Le Seuil.Find this resource:

Wiltschko, M. (2003). On the interpretability of tense on D and its consequences for Case Theory. Lingua, 113(7), 659–696.Find this resource:

Winford, D. (2006). The restructuring of tense/aspect systems in creole formation. In A. Deumert & S. Durrleman (eds.), Structure and variation in language contact (pp. 85–110). (Creole Language Library, 29). Amsterdam: Benjamins.Find this resource:

Winford, D. (2010). Substrate influence and universals in the emergence of contact Englishes: Re-evaluating the evidence. Presentation, June, at ECoLa 2010, University of Zurich.Find this resource:

Winford, D. (To appear). On the unity of contact phenomena: The case of imposition. To appear in C. Feral (ed.), In and out of Africa: Languages in question. Louvain: Peeters.Find this resource:

Wolfson, N. (1982). CHP: The conversational historical present in American English narrative. Dordrecht: Foris.Find this resource:


(1) . Following Comrie (1985), recognized as “absolute-relative,” in opposition to true or “pure” relative tenses such as those in Classical Arabic or in Latin participles. Hewson (this volume) analyzes these as combining primary tense with aspect, as many others have done. The pure relative tenses have received an aspectual analysis as well.

(2) . Not the same time t throughout the clauses of the sentence, though, but rather the same relationship between R and S. As far as I know, this was first stated by McGilvray (1974, p. 6), who wrote, “… ‘R’ need not designate exactly the same moment or interval in each clause [of a sentence]…. to require that it does in all [sentences] would make it difficult to analyze [George discovered that Harry had put a nickel in before the cop gave him a ticket] (and perhaps [If John had left quietly, Dick wouldn’t be in trouble]).”

(3) . A bibliography of works about and/or presenting Guillaume’s theory can be found at this date (December 17, 2010) at

(4) . Something of a related nature plays a role in Klein’s (1994, 1995) distinction between Reichenbach’s reference point and his own Topic Time (TT). For while R does not necessarily advance in either a sentence or a set of sentences expressing a temporal sequence of events, the TT does, thus maintaining the R point while providing narrative movement. We can identify the R-S relation with subjective, ascending time, and the TT-S relation with objective, descending time.

(5) . Also known as Fregean compositionality, after its putative originator, Gottlob Frege. First explicitly attributed to Frege by Carnap (1947), the principle is nowhere explicitly stated in Frege’s works, however. Pelletier (2001) argues that Frege did not, in fact, subscribe to it.

(6) . See for example Ogihara (1998).

(7) . Their example, from Google. A Google search on “I’ve had gone through” in fact returned almost 8.5 million examples, though not all are examples of the construction in question. But a large number of them are: “I appreciated the hard life I’ve had gone through …,” “O …,”ever the years I’ve had gone through several tanks …,” “I …,” “I now know I’ve had gone through many QJ[s],” “I …,” “I’ve had gone through one of my toughest week[s] here …,” “S …,” “Since age 3, I’ve had gone through 5 pacemakers,” “Anyone that has followed the case and what I’ve had gone through …,” “I …,” “I’ve had gone through that before …,” …,” etc., etc. This suggests that this construction is quite common in English and it is long overdue that serious attention be paid to it.

(8) . The history of the concepts of aspect and Aktionsart, and of the terminology associated with them, is quite complex. Short histories can be found in Binnick (1991, especially pp. 139–149), Kortmann (1991), and Młynarczyk (2004, Ch. 2).

(9) . “SUBINTERVAL verb phrases have the property that if they are the main verb phrase of a sentence which is true at some interval of time I, then the sentence is true at every subinterval of I including every moment of time in I. Examples of subinterval verb phrases are: walk, breathe, walk in the park, push a cart. NONSTATIVE, NONSUBINTERVAL verb phrases are verb phrases that are neither stative nor subinterval. Examples of such verb phrases are: die, walk to Rome, catch a fish, build a house. Since these verb phrases take the progressive form, they are nonstative. They are also nonsubinterval. Consider walk to Rome. If it took an hour to walk to Rome, one did not walk to Rome within the first thirty minutes of the hour” (Bennett & Partee, 1972/2004, p. 72).

(10) . A momentary or punctiliar action such as blink. The status of semelfactives as a type of Aktionsart is, as with so much else in aspectology, controversial.

(11) . Cf., retrieved September 10, 2010.

(12) . Various Latin grammars gloss the “frequentative” verbs as “to do x frequently”, “to do x much”, etc. Quite often there is little or no difference between the derived verb and its root verb. Thus the on-line Elementary Lewis-Short Dictionary glosses both cano and canto as “to produce melodious sounds, sound, sing, play.” However it does differentiate curro “run” from curso “run hither and thither, run constantly.”

(13) . The glosses are mine.

(14) . For a number of reasons, including the difficulty of fitting the perfect into a classification of the aspects such as Comrie’s, some scholars (Trager and Smith, 1951; Joos, 1964) have suggested that the perfect is not an aspect, but instead a “phase.”

(15) . Unaccusatives are verbs in intransitive constructions that are hypothesized to have underlying objects rather than subjects.

(16) . Unergatives are verbs in intransitive constructions that have subjects which are semantically agentive. Thus, leap is unergative while fall is unaccusative.

(17) ., retrieved November 14, 2010.