Syncretism (linguistics)

In linguistics, syncretism is the identity of form of distinct morphological forms of a word. This phenomenon is typical of fusional languages.

For example, in English, the nominative and accusative forms of you and it are the same, whereas he/him, she/her, etc., have different forms depending on grammatical case. In Latin, the nominative and vocative of third-declension nouns have the same form (e.g. rēx "king" is both nominative and vocative singular). Similarly, in German, the infinitive, first person plural present, and third person plural present of almost all verbs are identical in form (e.g. nehmen "to take", wir nehmen "we take", sie nehmen "they take").

Syncretism can arise through either phonological or morphological change. In the case of phonological change, forms that were originally distinct come to be pronounced identically, so that their distinctness is lost. Thus in the German case, the infinitive nehmen comes from Old High German neman, the first person plural nehmen comes from nemēm, and the third person plural nehmen comes from nemant. In the case of morphological change, one form simply stops being used and is replaced by the other: this is the case with the Latin example, where the nominative simply displaced the vocative in the third declension. Polyfunctionality, vagueness or syncretism? The concept of linguistic syncretism is a kind of controversial issue as not all morphologist agree as to which are good examples of syncretism and if those examples are straightforwardly appliable to any language or on the contrary, it might be idiosyncratic. Syncretism has also been called inflectional homonymy. However, since homonymy is normally applied to derivational morphology, trying to find a homogeneous definition for syncretism becomes a hard job. Divergence comes up with examples such as the word form "her" in English. Because of the many grammatical features that personal pronouns and determiners accumulate in just one morph, they are called in morphology cumulative expressions or portmanteau morphs. Thus, for example, one version of the lexeme SHE would be structured into the following grammatical morphemes: {3rd person}+{singular}+{feminine}+{object}+{ pronoun}. That is, only one morph for five grammatical morphemes. The other version of HER would be when it functions as a possessive determiner, so that the structure would be slightly different: {3rd person}+{singular}+{feminine}+{possessive}+{ determiner}. Thus, if we put them together, in both structures, there are three basic common grammatical morphemes that both versions share: person, number and gender, whereas they will differ in case and word subclass: SHE1: {3rd person}+{singular}+{feminine}+{object}+{ pronoun} SHE2: {3rd person}+{singular}+{feminine}+{possessive}+{ determiner} Is this a case of inflectional homonymy or syncretism? Syncretism is a matter of specification. It is an alternation of morphemes sharing a common core grammatical meaning, e.g. case alternation, number alternation, tense alternation or person alternation. The word form "her" features case alternation, i.e. possessive vs accusative. On the other hand, the form or morph realising the base morpheme in a lexical paradigm will naver be syncretic as the possible core meaning which might trigger alternation is underspecified. Haspelmath considers these cases of natural syncretism due to feature compatibility. According to Haspelmath, an example of natural syncretism would be regular nouns in English, which do not present case inflection but are case compatible, that is, they may be invariably used in nominative or accusative contexts. However, base morphemes never show a specific form for syncretic alternation. For example, the past tense forms of all verbs in English are neutral as for person and number specification. Thus, forms such as worked, met, bought or took could stand for 1Psing, 2Psing, 3Psing, 1Ppl, 2Ppl or 3Ppl. Thus, we may say that the past of work is worked, instead of saying: the 1Psing, 2Psing, 3Psing, 1Ppl, 2Ppl and 3Pp of work is worked in all the cases. We may say that worked is vague as person and number are concerned. However, it is specific as for tense and, together with met and bought would specify tense syncretism in two different ways, whereas worked added the past/past participle suffix to the base morpheme work, the latter modified the forms of their respective base morphemes (i.e. the root), meet and buy, to show past/past participle alternation. In both cases, we might assume an "archimorpheme" {PAST} which in some contexts is realised by the morpheme {past} and in other contexts it is realised by the morpheme {past participle}. This alternation would systematic for regular verbs in English as all regular verbs are syncretic in the past, and accidental for irregular verbs, as for example, the past tense for took will only be past, with the form taken realising the past participle morpheme. To assume that syncretism should always be specified by a marked form may be somewhat controversial, as there are forms such as "you", which shows no alternation for number or case, or "was" which features no alternation for person as they do not have a corresponding archimorpheme to specify that alternation. Thus "you" could either be nominative or accusative, and either singular or plural. Again, "you" would be a case of natural syncretism (Haspelmath 2002: 139). The rule would be: <<the second person pronoun in English is expressed by the form "you">> -with no reference to either number or case. This is reinforced by the fact that the form "you" is coincident with the base morpheme {YOU}. A slightly different case would be "was" as it is a suppletive form of the base morpheme {BE}. This form shows no alternation for number, as it is always singular, at least in standard English. However, it shows a clear alternation between 1P and 2P singular, though person does not seem to be specified, either. There is just one particular morpheme specification, the morpheme {PAST}, as a particular form has been chosen to express this grammatical feature, but no specification for either person or number. Concerning number, we might assume that is is an example of morphosyntactic blocking, as the plural is realised by the form "were", in the same way as irregular plurals of nouns are realised by root allomorphy in such words as mouse/mice, foot/feet. In these examples, the plural morph would block the regular realisation of the {PLURAL} morpheme. As stated above, irregular past forms in English do not specify number or person. The case of "was" could be similarly explained except for the fact that in the plural and second person singular this formed is blocked by "were". On the other hand, "were" underspecifies number and person. This could be explained the other way around. That is, the base allomorphs of the past of BE is "were" for all persons and numbers, except for 1P and 3P singular, where it is blocked by "was" in the indicative mood", as in the subjunctive mood, "were" would stand for all persons and numbers. This could be a case of "reverse" root allomorphy or "reverse" morphological blocking, as the Indicative mood is the unmarked mood as opposed to the unmarked subjunctive mood in BE, a unique case of subjunctive "inflection" in English, as subjunctive is never marked at least in the past, except for the form "were". Moreover, singular is the unmarked morpheme as opposed to the marked plural morpheme in both verbs and nouns. Moreover, the alternation of the word forms "was/were" could be considered a case of lexically-conditioned allomorphy. If we take the form "were" as the underlying root/morpheme as it is the only form which may stand for all persons and numbers, "was" would be a root allomorph of "were" elicited by the alternation of the grammatical morphemes 1P and 2P in the Indicative mood. The word form "are" would work in the same way as "were". However, in this case there is no blocking (except for, perhaps, the conversation clitic "aren't/ain't" for 1Psingular) and no syncretism.

Bibliography

Baerman, Matthew (2005). The syntax-morphology interface: a study of syncretism. Cambridge: Cambridge University Press. ISBN 0521821819. {{cite book}}: Unknown parameter |coauthors= ignored (|author= suggested) (help)