1. Defining the values

Map 102 depicts the number and identity of the arguments of a transitive clause which display person marking on the verb. “A” stands for the agentive argument and “P” for the patient argument. Five feature values are represented:

 

The lack of verbal person marking is illustrated in (1) by Bari (Nilotic, Nilo-Saharan; Sudan).

(1) Bari (Spagnolo 1933: 17) 

 

Person marking of only the A argument is exemplified in (2) from Kobon (Trans-New Guinea; Papua New Guinea).

(2) Kobon (Davies 1981b: 185) 

 

An example of person marking of only the P argument is given in (3) from Yawa (isolate; Papua, Indonesia).

(3) Yawa (Jones 1986a: 40) 

 

Person marking of the A or P but not of both is found, for instance, in Nocte (Naga, Tibeto-Burman; India).

(4) Nocte (Das Gupta 1971: 21) 

 

Observe that the person marking on the verb in (4a) is of the A, while in (4b) it is of the P. Whether it is the A or the P that is marked on the verb is dependent on which is higher on the person hierarchy 1 > 2 > 3.  If the higher ranking argument is a P rather than an A, the verb occurs with a special marker, h, in (4b), called an inverse marker.

The final value, person marking of both the A and P arguments, is shown in (5) from Tawala (Oceanic; Papua New Guinea).

(5) Tawala (Ezard 1997: 99) 

 

In most languages exhibiting verbal person marking, the person markers are affixes, attached to the lexical or auxiliary verb, as in the examples above. There are, however, also languages in which the person markers are not affixes but clitics. Clitics resemble affixes in forming a phonological unit with a word (their host), preceding the host (proclitics) or following it (enclitics). However, whereas affixes attach only to specific types of words (or stems), clitics attach to phrases and/or specialized syntactic positions. A common position for person clitics is the beginning of the verb phrase or verb complex. In Tinrin (Oceanic; New Caledonia), for example, the subject person markers are proclitic to the first element of the VP, which may be the verb (6a), a tense/aspect particle (6b), or a verbal modifier (6c).

(6) Tinrin (Osumi 1995: 215, 177, 182) 

 

Another common clitic position is after the first word or constituent in the utterance, as is the case in many southern Uto-Aztecan languages, including Southeastern Tepehuan (Uto-Aztecan; Durango, Mexico). We see below that the person marker is encliticized to a verb in (7a), to the end of a postpositional phrase in (7b) and to a question word in (7c). 

(7) Southeastern Tepehuan (Willett 1991: 194, 191) 

 

Though strictly speaking person markers such as those in Tinrin and Southeastern Tepehuan are not verbal person markers, they are treated here as such, provided that the verb is one of the constituents to which the person clitics may attach. By contrast, person markers which cannot be bound to the verb, i.e. independent person forms such as the subject markers in  Woleaian (Oceanic; Micronesia) in (8), or free-standing  combinations of person forms fused with tense as in Anejom (Central-Eastern Oceanic; Vanuatu) in (9), have been excluded.

(8) Woleaian (Sohn 1975: 93-94) 

 

(9) Anejom (Lynch 1982a: 119) 

 

Apart from affixal or clitic person marking, we occasionally find person marking via phonological changes in the stem. For example, in Misantla Totonac (Totonacan; Veracruz, Mexico), the second person singular A (and also S) marker is (with some verbs) marked by suppletion of the stem. This is the case with all verbs containing the roots /min/ ‘come’ and /an/ ‘go’. An example of the latter is given in (10).

(10) Misantla Totonac (MacKay 1999: 153) 

 

Such forms of person marking have also been taken into account here. 

Languages differ considerably in regard to the conditions under which they display verbal person marking. In some, verbal person markers are obligatory. In others they occur only in the absence of free nominal and pronominal arguments. As shown in (11), the latter is the case in Macushi (Carib; Brazil etc.) with respect to the person markers of both the A and the P.

(11) Macushi (Abbott 1991: 24) 

 

Rather less commonly, person markers occur in the presence of free pronouns but not nouns (e.g. Welsh, Luo) or alternatively in the presence of nouns but not free pronouns (e.g. Palauan). The occurrence of verbal person markers may also be conditioned by the location of the NP argument corresponding to the person marker (e.g. Apurinã, Cora, Paumari, Yagua), its animacy, definiteness or referential status (e.g. Daga, Lower Grand Valley Dani, Hua, Persian, Usan), the particular person involved (1st, 2nd, or 3rd) (e.g. Imbabura Quechua, Koasati, Panyjima, Parecis), as well as tense (e.g. Wolaytta), aspect (e.g. Awa Pit), mood (Evenki, Ika, Mountain Arapesh) and polarity (e.g. Estonian, Sentani). 

In classifying languages in terms of the number and identity of the arguments of a transitive clause manifesting person marking on the verb, most of the restrictions mentioned above have not been taken into account. Thus languages have been classified as exhibiting A, P or both A and P person marking on the basis of the maximal person marking that can occur in a single clause. Person markers which have not been included are: those found solely in dependent as opposed to independent clauses (e.g. Krongo), those occurring only in a minor mood such as the imperative (e.g. Nivkh) and those accompanying only a handful of verbs (e.g. Ingush).

It must be pointed out that some languages may lack markers of person on the verb for the A or the P, or even for both, but may have markers of number and/or gender. For instance, as shown in (12), in Archi (Nakh-Daghestanian; eastern Caucasus) there are prefixes on the verb (lexical or auxiliary) indicating the gender and number of the P (and also of the S) but not person. 

(12) Archi (Kibrik 1972: 124) 

 

That the verbal prefixes indeed mark only gender and not person is evidenced by the fact that while the prefix w- is used when the P is of male rational gender (12a) and the prefix d- when the P is of female rational gender (12b), the same w- prefix is used in (12a), (12c) and (12d), though the person of the P changes from third to second to first. To give another example, in Khanty (Ob-Ugrian, Uralic; Siberia) as well as various other Uralic languages (e.g. Hungarian, Nenets, Mordvin, Mansi, and Selkup) there are person suffixes for the A which are preceded by suffixes for the P. The latter, however, indicate only number and not person, as shown in (13).

(13) Khanty (Nikolaeva 1999: 24) 

 

Given that Map 102 is about the marking of person, markers of gender and/or number markers such as those in Archi and Ostyak have not been considered.

2. Geographical distribution

We see that verbal person marking is very common and widely attested cross-linguistically. In fact, only about a fifth of the languages in the sample lack any form of verbal person marking; such languages are found mainly in West Africa, the Caucasus, and South and Southeast Asia. Lack of verbal person marking is particularly rare in North America and New Guinea. Among the languages which manifest verbal person marking, around two thirds exhibit marking of both of the transitive arguments, the A and the P. This is the dominant pattern of person agreement in every area except Eurasia. It is particularly common in North America, New Guinea and Australia. In languages manifesting person marking of only one of the arguments of a transitive clause, this argument is typically the A (25%) rather than the P (7%). Verbal person marking of just the A is most common in Eurasia. It is characteristic of most of the Indo-European languages of Europe, the Uralic languages of Russia, Finland, Estonia and Hungary, the Turkic languages, and the Dravidian languages of India. It is also found in eastern Africa, especially among the Afro-Asiatic languages, in the north of South America and in New Guinea. By contrast, it is extremely rare in North America and there is no instance of it among the languages in the sample in Australia. Verbal person marking of just the P is an infrequent phenomenon everywhere. It is manifested mainly in Africa and South America but not at all in North America, Mesoamerica or Australia. In Africa it is found mainly among the Chadic languages, in a few Khoisan languages and sporadically in several groups of Niger-Congo languages.