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).

Subscriber: null; date: 13 August 2020

Poor (Two-Term) Case Systems: Limits of Neutralization

Abstract and Keywords

Two-term systems are of especial relevance to the diachrony of case systems: very often such systems represent either the last or the first stage of the existence of case in the language, and their study is important for understanding how case systems emerge and dissolve. To avoid confusion and aprioristic labels, the members of a bicasual system will be called Dir(ect) and Obl(ique). The label Dir is assigned to the case which coincides with the citation form of the noun. This article deals only with those systems where both cases are able to encode semantico-syntactic functions of verbal dependents (that is, relational cases). The article also discusses the areal and genetic distribution of two-term case systems, cross-linguistic trends in the functional organisation of two-term case systems, and their morphology and diachrony.

Keywords: two-term case systems, genetic distribution, functional organisation, morphology, diachrony, bicasual system, language

47.1 Introduction

Two-term (or bicasual) case systems are quite widespread in the world's languages (see section 47.2), but have not received enough attention from linguists. Nevertheless, they constitute an interesting phenomenon, whose study may be fruitful from various points of view: bicasual systems show what a ‘minimal’ case system may be like, which meanings of case markers may go together, how different patterns of argument encoding (accusative, ergative, etc.) may interact under extremely limited expressive possibilities (see sections 47.34). Two-term systems are of especial relevance to the diachrony of case systems: very often such systems represent either the last or the first stage of the existence of case in the language, and their study is important for our understanding of how case systems emerge and dissolve (see section 47.5). These are the issues which are going to be briefly discussed in this chapter.

The following two terminological remarks are in order. First, to avoid confusion and aprioristic labels, the members of a bicasual system will be called Dir(ect) (p. 687) and Obl(ique). The label Dir is assigned to the case which coincides with the citation form of the noun. Second, this chapter will be concerned only with those systems where both cases are able to encode semantico-syntactic functions of verbal dependents (i.e. relational cases); that is, systems where one of the two cases is a Genitive (as in Swedish) or a Vocative (as in literary Bulgarian) will not be discussed here.

47.2 Areal and genetic distribution of two-term case systems

Two-term case systems are attested in almost all major linguistic areas, although their distribution is far from being even. They are sporadically found in Europe, most notably in such already extinct languages as Old French and Old Provençal, but also in Romanian and in the dialects of other Balkan languages, in some Scandinavian dialects, and in modern English pronominals. In Asia bicasual systems are abundant in the Iranian, Dardic, and Nuristani languages, less in the Indo-Aryan languages (see Stilo, Chapter 48, for details), and are also attested in the NorthWest Caucasian languages (Adyghe and Kabardian). Such systems figure prominently in Africa, where they are found in almost all Berber languages, in the Ethiopian branch of the Semitic family, in many Cushitic languages (all belong to the Afroasiatic phylum), and in the Nilotic languages, which belong to the Nilo-Saharan phylum.

In the New World two-term case systems are not so common, probably due to the overall aversion of these languages towards dependent-marking. Here such systems are attested in the Salish language family, in some Uto-Aztecan languages, in Choktaw (a Muskogean language). In South America two-term case systems are found in a Bolivian isolate Movima, some Chibchan languages, and possibly in the Panoan language Matís, but it is probable that a closer investigation will reveal more such languages in that rather underdocumented region.

In the Pacific area two-term case systems are only sporadically attested, being found in Nias (an Austronesian language of Western Indonesia), in Yimas (a Papuan language of the New Guinea Highlands), in Maung (a Yiwaidjan language of Northern Australia, where case is restricted to independent pronominals), and in Aleut.

Thus, the languages with bicasual systems show great areal and genetic diversity, and it is no surprise that such case systems themselves exhibit considerable cross-linguistic variability. However, commonalities among two-term case systems found in the languages of the world are also quite noteworthy. Both similarities and differences among such systems will be discussed in the next sections.

(p. 688) 47.3 Functional properties of two-term case systems

When we look at a case system of a language, there are several questions we must answer in order to characterize it. The principal question concerns the range of meanings the cases are able to express. With respect to two-term case systems this question is rather divided into two:

  1. 1. Which semantico-syntactic functions are expressed by cases themselves (and not by adpositions)?

  2. 2. How are these functions distributed between Dir and Obl?

Thus, there are two major functional parameters of the typology of two-term case systems:

  1. 1. The ‘case zone’, i.e. the range of functions covered by the cases only, without use of adpositions.

  2. 2. The way the functions from the ‘case zone’ are distributed between the two cases.

Most languages with morphological cases possess also a more or less rich system of adpositions which express various meanings; bicasual systems are no exception, although there are languages with almost no adpositions among them (e.g. Salish and Yimas). We would expect that in a language with only two morphological cases and an array of adpositions the range of functions expressible by cases themselves must be rather limited, since in a two-term case system polysemy may easily lead to ambiguity. What we actually find, however, is that in the overwhelming majority of two-term case systems the ‘case zone’ is quite broad and usually includes, besides the core roles (S, A, and P; see Haspelmath, Chapter 33, for definitions), also such functions as Addressee/Recipient, Possessor, various locative and other circumstantial relations (cf. Stilo, Chapter 48). A typical system of this kind is found in Old French. Here Dir encodes S of intransitive verbs (see (1), Foulet 1970: 4), A of transitive verbs (see (2), Moignet 1976: 90), and nominal predicate (see (3), Foulet 1970: 8).

  1. (1) li chevalier-s s=en part.

    the(dir) refl=cl departs

    ‘The knight departs from there.’

  2. (2) il vit un home crucefié.

    he(dir) saw a( man( crucified(

    ‘He saw a crucified man.’

  3. (3) il est me-s pere.

    he(dir) is father(

    ‘He is my father.’

(p. 689) Obl, apart from P of transitive verbs (see (2) above), is used to encode the addressee with ditransitive verbs (see (4), Moignet 1976: 91), adnominal possessor (see (5), Foulet 1970: 14), goal of motion (see (6), Moignet 1976: 96), temporal interval (see (7), Moignet 1976: 95), manner of action (see (8), Foulet 1970: 32):

  1. (4) dites le roi que ….

    say:imp.2pl the( king( that

    ‘Tell the king that …’

  2. (5) la niece le duc

    the niece the( duke(

    ‘the niece of the duke’

  3. (6) droit sentier qui cele part

    direct( road( that( this( place(

    le menast.

    he( would.lead

    ‘[He could not find] a direct road that would lead him to that place.’

  4. (7) Erec dormi po cele nuit.

    Eric( slept little this( night(

    ‘Eric slept a little this night.’

  5. (8) s'=en part le-s gran-z galop-s.

    refl=cl departs

    ‘[And the knight] departs in great gallop.’

Other functions found in two-term case systems include Instrument (see (9) from Squamish, Salish, Kuipers 1967: 169), Location (see (10) from Yimas, Foley 1991: 166), Comitative (see (11) from Movima, Haude 2006: 282).

  1. (9) na=λič’itas ta=smic t=ta=λač'tn.

    asp=cut:3sg.a/3sg.p art=meat obl=art=knife

    ‘He cut the meat with a knife.’

  2. (10) ŋaŋk-ŋan ama-na-irm-n.

    grass-obl 1sg-def-stand-pres

    ‘I am standing in the grass.’

  3. (11) kide: da’ kaykay jayna n=us alwaja=‘ne.

    they dur eat:rdp now obl=art spouse=3sg.f

    ‘They are eating now with her husband.’

Functionally rich systems like the Old French one are very common; by contrast, ‘narrow’ systems, where the case zone is limited just to core cases, or includes only one or two peripheral functions, are rare (cf. the Berber languages, Aleut, and Wakhi, an Iranian language of Pamir). Such a cross-linguistic distribution of (p. 690) ‘broad’ vs. ‘narrow’ two-term case systems implies that languages perfectly tolerate extended polysemy of case markers. This is due to the general tendency of highly grammaticalized case markers to encode particular functions only with those nominals which are ‘natural’ with these meanings (see Aristar 1997). For instance, Obl may be interpreted as ‘locative’ with names of locations, as ‘instrumental’ with names of instruments, and as ‘dative’ or ‘comitative’ with animate nominals.

Let us turn to the second parameter, the distribution of functions among the two cases. If we first look at the encoding of core grammatical relations, in the languages with bicasual systems we will find all possible kinds of marking. The nominative-accusative marking is most common (e.g. Old French, Uto-Aztecan, Berber, Nilotic, Amharic, Persian, etc.); next comes the neutral encoding (Salish, Yimas, Movima, Aleut). The ergative-absolutive marking is dominant (which does not mean ‘unique’) in Adyghe and Kabardian, Nias, Matís, and Päri, a Nilotic language, but occurs as an option in Aleut and in many Indo-Iranian languages. The S/A or S/P participant is usually encoded by Dir, but there are notable exceptions, cf. examples (12) and (13) from Kabyle (Berber, nominative-accusative, Chaker 1983: 276, 279), and (14) from Nias (ergative-absolutive, Brown 2001: 94):

  1. (12) fγ-n y-rgaz-n.


    ‘The men left.’

  2. (13) y-wt aqšiš-ni w-rgaz-im.

    3sg-hit dir:boy-this

    ‘Your husband hit this boy.’

  3. (14) me mofanö ya, la-roro ya niha fefu.

    when left he(obl) 3sg-follow he(obl) (dir)person all

    ‘When he left, everyone followed him.’

In Kabyle, as well as in Nilotic, Cushitic, and Muskogean languages, it is the S/A relation which is marked by Obl, not the P. Similarly, in Nias the A is encoded by Dir, whereas the S/P participant receives morphological marking. The rationale of such systems lies not in the alleged ‘unmarkedness’ of the S relation (cf. Comrie 1978), but in general markedness principles (cf. Givón 1995; Haspelmath 2006): among the two cases in a minimal system it is the ‘default’ case used to encode many different functions which remains the unmarked member of the morphological opposition. By contrast, the case whose only function is to encode the ‘subject’ (S/A) argument, is both functionally and formally marked. The next question is why the ‘subject’ function is not encoded by the ‘default’ Direct case in these languages. The answer probably lies in the realm of information structure: it is usually only the non-topicalized and thus functionally marked ‘subject’ participant which receives Obl encoding in these languages, see the following examples from Tachelhit (Berber, Galand 1964: 34, 40):

  1. (p. 691) (15) ikrz u-rgaz igr.

    cultivated obl-man dir:field

    ‘[It was] the man [who] cultivated the field.’

  2. (16) a-rgaz ikrz igr.

    dir-man cultivated dir:field

    ‘[As for] the man [he] cultivated the field.’

Therefore, such ‘marked nominative’ systems are in fact functionally well motivated (see König, Chapter 35, and König 2006 for a more detailed discussion).

Among the two-term case systems ‘split’ case marking is very common; in the Indo-Iranian group, where there is both a tense-aspect split in the marking of A, and definiteness/animacy split in the marking of P, up to four constructions (neutral, accusative, ergative, and double-oblique) may co-exist in a single language (cf. Stilo, Chapter 48). A clause-type split is found in the Uto-Aztecan languages, where the S/A participant is marked by Obl in subordinate clauses (see (17) from Yaqui, Lindenfeld 1973: 103).

  1. (17) na=a biča ke hu-ka usi-ta čuɁu-ta kipwe- Ɂu

    1sg:dir=3sg see that this-obl child-obl dog-obl. have-prf

    ‘I see that this child has a dog.’

If we turn to the general patterns of the functional organization of two-term case systems, we may find two principal patterns of distribution of meanings between the cases:

  1. 1. ‘Dividing’ systems, where all peripheral functions are attributed to a single case (usually Obl), which may also have a core function;

  2. 2. ‘Distributing’ systems, where both cases have core as well as peripheral functions.

‘Dividing’ systems are by far the most common (compare Old French above and Table 47.1), while the genuine ‘distributing’ systems occur only in some languages of the Pamir and Hindukush region, for example in the Nuristani language Kati (see Table 47.2, Èdel'man 1983: 60–1).

Table 47.1. Functions of cases in Old French


S, A, Pred


P; Addressee, Possessor, Locative, Goal, Temporal, Manner

Such an uneven distribution of the two types of two-term case systems is probably due to the general tendency of cases to encode ‘natural’ classes of functions, such as core vs. peripheral or S/A vs. all others. By contrast, in the ‘distributing’ systems such as that of Kati the only rationale for the ‘division of labour’ between the cases is their diachronic origin: the Indo-Iranian Dir stems from the collapse (p. 692) of older Nominative and Accusative, while Obl derives from Genitive-Dative. In the ‘distributing’ systems the two cases retain the functions which belonged to the different cases they originate from. It is noteworthy that the majority of the Indo-Iranian languages must have undergone a functional change and redistributed the functions of cases, so that now their two-term case systems are of a genuinely ‘dividing’ type.

Table 47.2. Functions of cases in Kati


S, A, P; Goal, Locative


A in the past tenses, definite P; Addressee, Possessor

To summarize, from the point of view of the semantic content of a case system, bicasual systems may be characterized by the following properties:

  1. 1. The cases usually cover a broad range of meanings, including both core grammatical relations and peripheral functions (locative, temporal, manner, etc.).

  2. 2. The markedness relations between Dir and Obl tend to iconically reflect the functional load of these forms: the case with a greater variety of uses and a non-restricted distribution is usually the morphologically unmarked Dir, even though the ‘subject’ relation may be encoded by the other case.

  3. 3. The distribution of functions between cases more often than not follows the pattern where all peripheral functions are subsumed under one of the cases only.

47.4 Morphological properties of two-term case systems

In the previous section the cross-linguistic trends in the functional organization of two-term case systems were surveyed. Now let us consider the morphological make-up of such systems. Here we find that typologically rare and unusual patterns appear with a frequency greater than average. This concerns both form and position of case exponents attested in bicasual systems. The most frequent type of marker used in such a system is a bound affix, but there are deviations from this prototype in both directions. Thus, in Halkomelem and some other Salish languages, in Amharic, and Persian, Obl is a clitic, cf. example (18) from Amharic where the case marker – әn behaves as a second-position clitic attaching to the preposed adjective rather than to the head noun (Leslau 1995: 184).

  1. (p. 693) (18) wәšša=w tәllәq=u=n bäqlo näkkäsä.

    dog=art big=art=obl mule bit

    ‘The dog bit the big mule.’

By contrast, in Nias Obl is normally realized as a morphophonological process, that is, ‘mutation’ of the initial consonant of the stem, cf. Table 47.3 (Brown 2001: 39–40).

Table 47.3. Mutation as case exponence in Nias















Morphophonological alternations function as case exponents, usually alongside affixes, also in Old French and Old Provençal, in many Indo-Iranian and Afroasiatic languages. Finally, in Nilotic and Cushitic languages the primary and more often than not the only exponence of case is tone (see Bennett 1974), see the paradigms from Maasai in Table 47.4 (Tucker and Bryan 1966: 459).

Table 47.4. Tone as case exponence in Maasai


















If we now look at the position of case markers with respect to the stem, we find that the well-known ‘suffixing preference’ (see e.g. Hawkins and Cutler 1988) is in two-term case systems less prominent than in the languages of the world in general. According to Dryer (2005b), preposed case markers are found in less than 10 per cent of the languages with cases. However, among the languages with bicasual systems prefixal case markers are found in about 30 per cent of the linguistic groups, namely in Berber and Salish languages, Nias, and Movima. It is not at all obvious how these figures are to be interpreted, but they are nevertheless quite significant.

Turning to more complex issues, nominal paradigmatic structures observed in two-term case systems are often non-trivial. Certainly, the most common option is a separate exponent of (usually only Obl) case, invariable across different nominals; however, various deviations from this simple structure are attested. First of all, number and sometimes gender may be encoded cumulatively with case, as for example in the Indo-Iranian languages. Moreover, number may be expressed (p. 694) separately, the case exponent being nevertheless sensitive to it, cf. the Khowar (Dardic, Èdel'man 1983: 212) paradigms in Table 47.5.

Table 47.5. Case and number exponents in Khowar

















Various types of neutralization of categories are found, too. It is certainly common to have no case distinction in the plural, as for example in Yaqui and Aleut, but some languages (many Indo-Iranian throughout all nominals, as well as Old French in the subset of demonstratives) neutralize number in the Direct case, cf. the Kati (Nuristani, Èdel'man 1983: 60) paradigm in Table 47.6.

Table 47.6. Nominal paradigms in Kati















In the Indo-European two-term case systems more ‘exotic’ patterns of syncretism are also found, such as the identity of Oblique Singular and Direct Plural, observed in many Indo-Iranian languages, cf. Table 47.7 with Pashto paradigms (Skjærvø 1989a: 390).

Table 47.7. Nominal paradigm in Pashto





 Poor (Two-Term) Case SystemsLimits of Neutralization


 Poor (Two-Term) Case SystemsLimits of Neutralization



 Poor (Two-Term) Case SystemsLimits of Neutralization


 Poor (Two-Term) Case SystemsLimits of Neutralization


In Old French not only Oblique Singular and Direct Plural fall together, but quite often Direct Singular and Oblique Plural, too; this has led to a situation when four paradigmatic cells are covered with only one overt affix, cf. Table 47.8 (Pope 1934: 311). (p. 695)

Table 47.8. Nominal paradigm in Old French










To conclude this section, we may observe that the morphological make-up of bicasual systems has some peculiar characteristics which are seldom or never attested in richer case systems. Minimal systems are more prone to prefixal or non-concatenative case marking, as well as to unexpected patterns of paradigmatic neutralization.

47.5 Diachronic issues

There are two main questions concerning the diachrony of two-term case systems:

  1. 1. What are their possible diachronic sources?

  2. 2. What happens to them in the course of their history?

These issues will be briefly discussed in this section.

For the majority of the languages discussed in this paper, no written sources concerning their history are available. Therefore, any hypotheses about the origins of two-term case systems outside the Indo-European family are rather tentative. It is possible to discern the following two processes leading to a two-term case system: (i) reduction of a richer case system (as in the Indo-European languages); (ii) grammaticalization from other types of markers, such as adpositions (as in Salish, Movima, and probably Yimas) or demonstratives/articles (as in Berber, see Chaker 1988, and Ethiopian Semitic). The origins of tonal case systems of the Nilotic languages remain obscure.

The process of disintegration of polycasual systems into bicasual ones is well documented both for the Romance languages (e.g. Pope 1934) and for the Indo-Iranian languages (e.g. Kerimova and Rastorgueva 1975). It seems that various factors were interacting in the course of this development. Besides the obvious phonological erosion of case endings, syntactico-semantic processes were also of the utmost importance. The overlapping of the range of uses of the original cases led to their becoming interchangeable in many contexts, and to a decrease in the number of (p. 696) grammatical oppositions. Numerous case syncretisms existing already in the older polycasual systems also facilitated the functional and morphological merger of originally different cases.

Of especial interest in this connection is the situation observed in Sogdian (Middle Eastern Iranian, Sims-Williams 1982), where an older system of five cases co-existed with a newer two-term case system. Such a rare situation became possible because of the so-called ‘rhythmical law’ (Tedesco 1926), which caused reduction of final syllables of the so-called ‘heavy stems’, with the ‘light stems’ remaining intact, cf. paradigms of masculine nouns of both types in the singular in Table 47.9 (Sims-Williams 1982: 67, 68).

Table 47.9. Nominal inflection in Sogdian (masculine, singular)

‘people’ (light)

‘day’ (heavy)















However, the rhythmical law was not the only reason for the emergence of the bicasual subsystem in Sogdian; the two-term Dir ~ Obl distinction was observed throughout the feminine gender and in the plural of all types of nouns, where it appeared probably even prior to the operation of the rhythmical law. Thus the older polycasual system was a marked option restricted to a limited subset of nominals, in contrast to a newer two-term case system appearing elsewhere.

In the process of reduction of an older polycasual system into a bicasual one several older cases fall together thus forming a new case with a broader range of meanings. The resulting set of functions is not necessarily the simple unification of the uses of the predecessors of the new case, but the way the older system collapsed into a bicasual one crucially affects the structure of the latter. We have already seen (section 47.3) that the ‘distributing’ systems in the languages of the Hindukush-Pamir region reflect an earlier stage of development and retain the uses of older cases, while the majority of the ‘dividing’ systems in the Indo-Iranian languages must have redistributed the original case functions. If we compare the Indo-Iranian and the Old Romance two-term case systems, we see that the neutral alignment attested in the former and the predominant accusative alignment of the latter has clear diachronic origins. Indeed, during the disintegration of the Ancient Indo-Iranian polycasual system the original Nominative and Accusative fell together, (p. 697) whereas in the course of the decline of the Latin case system the distinction between the Nominative and all other cases was usually retained (see e.g. Plank 1979).

Let us now turn to the fate of two-term case systems. Besides persisting for a long period without major changes (as probably is the case in the Nilotic languages), there are two major pathways of change:

  1. 1. A bicasual system may serve as a base for a newly grammaticalized polycasual system.

  2. 2. A bicasual system may disintegrate thus leaving the language without the case category altogether.

The first scenario is observed in the Indo-Aryan languages, where the primary function of the older inflectional two-term case system is to host a whole series of postpositions or even already bound case affixes (see Masica 1991: 238–48 for details). The postpositions usually attach to the Oblique case, and the new declensional system thus has two stems, one for the Direct (or Nominative) case, and another for all or most peripheral cases. The older Oblique itself in some languages may be used in isolation, thus becoming a new case, e.g. Accusative (as in Romani, Table 47.10, Wentzel 1980: 72), paradigmatically opposed to other cases formed upon its base, or, in others, functions only as the bound oblique stem, as in Gujarati (Table 47.11, Saveljeva 1965: 24–5).

The postpositional origin of such polycasual systems reveals itself in the behaviour of adjectives, which usually show up in a common Oblique form with all non-Nominative heads, cf. example (19) from Gujarati (Saveljeva 1965: 28).

  1.  Poor (Two-Term) Case SystemsLimits of Neutralization
     Poor (Two-Term) Case SystemsLimits of Neutralization

    (19) glel svapn-o-m

    ‘in delirious dreams’

Finally, let us look at the situations when a two-term case system disintegrates. As in the fall of a rich case system, many factors play a role here. Sometimes it seems that Dir extends its usage, simply ousting Obl from its original contexts. Thus, in some Norwegian dialects Obl with nouns is used in the following functions: (i) as (p. 698) the Recipient of ditransitives; (ii) as a ‘quirky’ object of some non-canonically transitive bivalent predicates; (iii) as the object of some prepositions with a locational meaning (‘live in the town’), contrasting with the allative meaning of Dir (‘go to the town’); (iv) as a governed object of other prepositions. It appears (Sandøy 1996: 134) that Dir is taking over the more syntactic uses of Obl, viz. (ii) and (iv), while the latter persists in those contexts where it is semantically opposed to Dir, viz. (i) and (iii). Thus the range of functions Obl can encode gradually shrinks, and not in an unprincipled way.

Table 47.10. Nominal declension in Romani (North-Russian dialect)














Table 47.11. Nominal declension in Gujarati






 Poor (Two-Term) Case SystemsLimits of Neutralization









 Poor (Two-Term) Case SystemsLimits of Neutralization

A more complex situation was observed in late Medieval French (fourteenth-fifteenth centuries, see Laubscher 1921). Here the system was inherently unstable due to the typologically unusual markedness of the Direct case, coupled with its rather restricted range of functions, as well as the loss of phonologically overt markers. The less marked and more frequently used Obl began to gradually take over the uses of Dir, but there was a short period of free variation. As a result, in most cases Obl was retained and Dir lost, but sometimes they gave rise to different lexemes (e.g. sire 〈 Dir in vocative function vs. seigneur 〈 Obl), and in others it was Dir rather than Obl which persisted (e.g. père ‘father’).

47.6 Conclusions

In the preceding sections we have discussed various properties of two-term case systems, both functional and formal. Let us briefly review the principal points.

First of all, despite the seeming poverty of expressive means, two-term case systems more often than not cover quite a wide range of different semantic roles, not only the core grammatical relations, but also a more or less rich array of peripheral and circumstantial roles; ‘poor’ two-term case systems covering only the core relations are rather rare.

(p. 699) Second, markedness relations between cases in two-term case systems are usually driven not only by the typologically stable associations of marked expression with more ‘peripheral’ semantic roles and unmarked expression with more ‘central’ semantic roles, but also by the functional load of the cases. ‘Marked nominative’ systems, where the S/A relation is encoded by a marked case, while other semantic roles fall under the unmarked case, are quite widespread here.

Third, in two-term case systems such typologically rare case exponents as prefixes and proclitics or tonal modification are attested. Also, ‘minimal’ systems allow for paradigmatic structures which are usually not found in richer case systems, for example the Old French X-like neutralization of case and number.

Finally, from a diachronic perspective, two-term case systems constitute an important stage in the development of case in a language. They may be either the last stage of the reduction of an older polycasual system, as in the Romance and Iranian languages, or serve as the basis for a newer case system emerging via grammaticalization of adpositions usually attaching to the Oblique case as the stem.