C2: "Perfect Mapping: Typological Similarities and Variations of the Perfect across Languages" (Thomas Jügel)

Mon 11.15-12.45, Tue 14.15-15.45, Wed 16.15-17.45, Fri 9.15-10.45
Room: 1.01


The verbal category ,perfect‘ is difficult to define cross-linguistically. Tense, aspect, and aktionsart play a role and languages do not treat these functions homogeneously. In consequence, the functional range of expressions that are called ‚perfect‘ differs as do the linguistic definitions of tense, aspect, and aktionsart, because they depend on the sample of languages the terminology is applied to. On the one hand, language specific definitions cannot be applied cross-linguistically, and on the other hand, universal approaches face the challenging task of covering innumerable varieties.

The key is to decompose the functional range of the perfect and to separate the core function from interpretations (semantic composite effects), i.e., to set up a map of the semantics and functions of the perfect. We will approach this task by concentrating on the following four topics:

Grammaticalisation paths of perfects (focus: Iranian languages)

According to Heine (1993: 68), the perfect is often linked in the following path of grammaticalisation: „completive/resultative > perfect > perfective > past > irrealis“. The development of resultatives to past can be tracked in Persian. The Old Iranian resultative construction (as it appears in Avestan) functions as a perfect in Old Persian. Its perfective use becomes apparent in Middle Persian, while New Persian shows a stronger tense interpretation. A few aspectual uses remain, though.

Analytic perfect constructions (focus: Romance languages)

Western Europe is famous for its analytic perfect construction with the auxiliaries to be and to have. Dahl (1995: 365) claims that the ‚have-perfect‘ is almost exclusively found in Europe. However, such constructions can be found in a variety of languages of the past and present (e.g., Hittite, Sogdian, etc.). Nevertheless, Romance languages are good examples for investigating the historical development. The common claim that a possessive construction was at the root of this analytic construction (Benveniste 1952, Vincent 1982, and many others) is challenged by the evidence of the data (cf. Thielmann 1885) and by theoretical considerations. For instance, Jacob (1998: 113ff.) states that the experiencer or benefactive characteristics of logical subjects in perfect, modal, and stative constructions are the reason for the frequent use of possessive auxiliaries in such constructions and he rejects the literal possessive interpretation of perfects.

Modal use of perfects (focus: Turkic languages)

Modal use, like the inferential interpretation of Turkish perfects, is commonly derived from a resultative perfect, the event of which was not witnessed by the speaker (cf. Johanson 2000: 63f.). Then again, factivity is a feature of the New High German perfect (cf. Thieroff 1994: 100). Modal interpretation can also be expressed by imperfective forms (e.g., the imperfect for irrealis and the progressive for factivity in New Persian). So the question of whether such modal usage is part of the semantics of the perfect or context-sensitive interpretation arises.

Perfective vs. perfect (focus: Slavic languages)

Many scholars consider Russian a proto-typical aspect language with a perfective and an imperfective verb for any verbal lemma and its forms. The formation of aspect is closely connected with a rich aktionsart system. However, the Russian tense system is very basic: past, present, and an analytic future. How does Russian express a perfect? Is equation of perfective and perfect possible for a subdomain of the perfective?


Lecturer: Thomas Jügel, UMR Mondes iranien et indien, INaLCO, Paris