|Home||Feature Inventory||Related Issues||Search||Contact|
Person is a deictic category, interpreted relative to the speaker, encoding the participants in a speech situation. The cognitive foundation of the feature of person reflects the basic structure of a speech act and distinguishes the following speech act participants: the speaker and the addressee, and what is spoken about (cf. Benveniste 1966).
Reference to the participants in a speech act can be expressed linguistically in various ways. In English, in certain contexts it is possible to refer both to oneself and to one's addressee using common nominal phrases, as in Mummy will help you, or Would Your Honour like to see the evidence now?. However, these nominal phrases also have a non-deictic function. On the other hand, elements such as I or you, as well as markers of agreement in person, are used exclusively for participant deixis: they are specialised 'shifters' (Jespersen 1922:123; cited in Cysouw 2003:5) which normally have no other usage besides shifting their reference to different extralinguistic entities particular to each communicational setting.
Such specialised shifters which are used for reference to participants in the speech act are often referred to as 'person markers'. The set of which they are part forms a paradigm which is traditionally analysed as one collapsed dimension together with number. Some linguists, however, do not give the combination person-number special status, but treat person as an individual attribute, separate from number. Cysouw (2003:101 ft. 2) mentions the following in this group: Jespersen (1924:212-215), Lyons (1968:276-281; 1977:636-646) and Croft (1990:145-150). Cysouw (2003) also separates the two features. According to this view, the paradigm of person includes: 'speaker' ('1'), 'addressee' ('2'), and 'other' ('non-participant'), and based on these distinctions, groups of participants can be formed (see below in §4). Groups of participants consist of more than one participant and are thus necessarily semantically plural.
The biggest analytical problem for the category of person arises when the speaker ('1') is involved in the participant group. The most common distinction within the range of possibilities is first person inclusive which is understood as speaker and addressee ('1+2'). Under such a system the first person exclusive is understood as including the speaker but excluding the addressee. However, other possible combinations of participants may involve the issue of how one interprets number marking for such groups. Furthermore, number values can additionally impose restrictions ('two', 'three', 'small number') on the number of participants in the groups. Therefore, even though we accept that the number paradigm can cross-cut the person paradigm (Corbett 2000:64-66, 83-87), the interaction of person and number is not straightforward to capture.
The inclusive/exclusive distinction (applied typically with regard to first person) is one of the important distinctions that have been identified for the category of person. Another one is the proximate/obviative distinction (applied typically with regard to third person). While the inclusive/exclusive distinction is typically defined as expressing the inclusion of the addressee with the first person, the proximate/obviative distintion regards the degree of remoteness of the non-participant.
It is worth considering the possibility that the inclusive/exclusive, proximate/obviative and other distinctions that are associated with the person feature could be treated as separate concepts which intersect with person marking. For example, the proximate/obviative distinction is about the degree of remoteness from the speaker when talking about non-participants. Equally, the inclusive/exclusive distinction is about inclusion of the addresse, a participant which is proximate to the speaker. It is worth noting that the terms 'inclusive' and 'exclusive' have also been used in relation to the second person in Abkhaz, where there is a distinction between 'you-excluding-them' and 'you-including-them' (Cysouw 2003:75 and references therein). Given the standard definition of inclusive/exclusive as either involving the addressee or not, the extension of this concept to the second person would involve a logical contradiction. However, there is another possibility, namely that there is a general concept involving the degree of remoteness relative to a speech act participant. First person inclusive and first person exclusive could be understood as the intersection of the first person with this generalised concept, and examples of the type found in Abkhaz could be understood as the intersection of the second person with this general concept. Finally, the proximate/obviative could be understood as the intersection of the third person with this general concept (thanks to Dunstan Brown for a very fruitful discussion of these issues).
The category of person in a language has to be expressed through morphology in order to be regarded a feature, whether morphosyntactic or morphosemantic (see also §3 below).
Investigation of morphosyntactic expressions of person reveals that languages with personal inflection differ greatly with respect to which and how many of the available person values are expressed in a single predication. The choice of the expression of the person value may be determined by the relative position of the participant in a person hierarchy. One possible hierarchy of this type (part of a more extended animacy hierarchy) has been formulated as follows (Silverstein 1976; Comrie 1981):
1st/2nd person > 3rd person
A person hierarchy captures the fact that participants can be referred to by person values independently of their semantic or syntactic status (Helmbrecht 1996:129). However, the person-based reference to arguments in a clause can also be controlled by syntactic functions:
Subject > Direct Object > Indirect Object
or semantic roles:
Agent > Recipient/Experiencer > Patient
(cf. Givon 1976:152; Croft 1988:162ff). The hierarchy of syntactic functions captures the observation that, in languages with grammaticalised functions of subject and object and syntactically controlled person agreement, the predicate is the most likely to agree in person with the subject, before we find agreement with the direct object or indirect object. On the other hand, the hierarchy of semantic roles captures the generalisation that, if person agreement is controlled by semantic functions of the arguments, the predicate is the most likely to agree with the agent, followed by the recipient/experiencer (which is frequently sentient, therefore it will tend to be human or at least animate), followed by the patient (which can frequently be inanimate and indefinite). Moreover, it is possible that more than one hierarchy can be in use to control person agreement in one language (e.g. Akhvakh, East Caucasian, which predominantly uses a nominative-accusative syntactic strategy to code first person transitive agent, but with some less prototypical transitive verbs the first person marking is controlled by the semantic role of the ergatively marked experiencer; Helmbrecht 1996:137). For more detailed information about various aspects of person marking, see Siewierska (2004). Furthermore, Siewierska (2005a-d) gives an overview of verbal person marking and verbal person marking alignment types.
The controllers of agreement in person are linguistic elements that express syntactic arguments - these are typically nouns or pronouns, but may also be pronominal affixes (see e.g. Dryer 2005). For detailed discussion of pronominal affixes, and the diagnostics for distinguishing pronominal affixes from agreement markers, see Corbett (2003), also summarised in Corbett (2006:99-112). Note that, on this account, pro-drop phenomena (where, in terms of syntax, it may be arguable whether a given element represents an argument or does not fill the argument slot) are not treated uniformly, but need to be analysed on a language-by-language basis, according to the criteria for identifying agreement in person.
The category of person has often been assumed to be universal (Forchheimer 1953:1; Greenberg 1963:31,96; Benveniste 1971:225; Wierzbicka 1976, 1996; Zwicky 1977:715; Ingram 1978), and the claims have varied from a reference to a rather vague 'expression of person' (Benveniste) or 'the system of person' (Forchheimer) to specific remarks about the universal existence of 'distinct first and second singular independent pronouns' (Greenberg), 'pronominal categories involving at least three persons and two numbers' (Greenberg), or the 'morphosyntactic categorisation of person' (Zwicky).
From the point of view adopted here, the (cognitive) category of person exists in a language if it is possible to make a distinction between at least two of the basic participants in a speech act. This is achieved, for example, by allowing self-reference or reference to the addressee. Such reference can be made with the conventional use of any type of noun, or by using some special words that lexicalise the meanings of 'speaker (1)' and 'addressee (2)'. However, the morphosyntactic feature of person reflects the grammaticalisation of the category of person in the language: specifically, the participation of person values in agreement. The existence of personal pronouns, without any influence of the category of person on decisions regarding agreement, is not sufficient to posit the person feature for the language, since the pronouns may be lexicalised meanings for the participants of the speech act.
Thus, person as a morphosyntactic feature is a feature of agreement. When it is found on controllers of agreement, it is an inherent feature, and when it is found on targets of agreement - it is a contextual feature (see the 'Feature Inventory' page for clarification of this distinction). The controllers of agreement in person are linguistic elements that express syntactic arguments (though see a note above in §2 about pro-drop).
In languages where free pronouns have a generic pronominal root, typically invariant across all person-number values, with person affixes attached (Siewierska 2004:19), the pronouns have a 'selected' feature specification for person, since the person value is selected from a range of options. "Etymologically, the generic pronominal root is often the word for person, body, self or the verb 'to be' or 'exist'. It may, however, be some other form. For instance, in Warekena, an Arawakan language of Brazil, in the case of the first- and second-person and third-person plural, the person prefixes are attached to the emphatic root -ya and in the case of the other third-person categories to forms cognate with demonstratives" (Siewierska 2004:19). In languages where pronouns do not inflect for person in this way, the person specification can be regarded as 'fixed' (lexically supplied).
It is assumed that noun phrases have inherent third person value, and that they get this specification by default (Corbett 2006:132). There are interesting cases in which this default can be overriden, involving syntactic contexts similar to apposition ('[we] women are...', etc., and the use of the vocative case when the person specification in these contexts is second singular or plural).
As was already mentioned in §1, the paradigm of person marking is traditionally analysed as one collapsed dimension together with number. Cysouw (2003) list linguists who have separated the two dimensions, and suggests that the paradigm of person includes: 'speaker' ('1'), 'addressee' ('2'), and 'other' ('non-participant'); based on these distinctions, groups of participants can be formed. Groups of participants consist of more than one participant and are thus necessarily semantically plural. Seven different group combinations of the three basic singular participants are theoretically possible, but only five of them seem to be attested: 1+2 (minimal inclusive), 1+3 (exclusive), 1+2+3 (augmented inclusive), 2+3 (second person plural), and 3+3 (third person plural) (N.B. According to Cysouw, no grammatical form has been attested for the following, although the uses are attested: 1+1 'choral we', 2+2 'only present audience'). Cysouw argues that these groups, together with the three singular participants, form the basic paradigm for the typological classification of person which is unmarked for number, because the number of participants in the groups is not relevant. (Additionally, dual, trial and other number values can impose further restrictions - 'two', 'three', 'small number' - on the number of participants in the groups, and in this way the number paradigm can cross-cut the person paradigm.)
However, it is not obvious that Cysouw's method does really separate person from number. The biggest problem arises when the speaker ('1') is involved in a participant group. It is widely accepted that first person inclusive is usually understood as speaker and addressee ('1+2'). Under such a system the first person exclusive is understood as including the speaker but excluding the addressee. However, other possible combinations of participants do involve the issue of how one interprets number marking for such groups (see below).
What follows is an attempt at constructing a value inventory for the feature person, with a clarification of the problem of the interaction of person and number (with thanks to Dunstan Brown for discussion that has clarified this, and for the contribution of the definitions):
Logophorics and long distance reflexives are considered to be subclasses of third person. Finally, there are languages which distinguish more than two degrees of remoteness from the speaker, but usually only in the pronouns (with no further distinction than two degrees of remoteness in agreement).
Crosslinguistically, there are many different patterns of syncretism between the values of the full person paradigm, especially when considered jointly with the cross-cutting number paradigm (Cysouw 2003; Siewierska 2004; Baerman, Brown & Corbett 2005).
Personal pronouns can be morphosyntactically odd in different ways. In some languages they may have unexpected inflectional properties. For example, the whole class of pronouns can have a particular grammatical gender irrespective of the fact that they may refer to persons of different genders. This has been reported for Jarawara (a dialect of the Madi language of the Arawá family, spoken in southern Amazonia), in which all pronouns take feminine agreements, irrespective of the sex of the referents (Dixon 1995:265, 290). Another interesting instance comes from Burmeso (a language isolate spoken in northern Irian Jaya), where the first singular pronoun takes feminine agreement and the second singular takes masculine (Donohue 2001:100-101).
Barasano (an eastern Tucanoan language spoken in Colombia) shows a different, curious interrelation between person and gender: subject agreement markers (suffixes) in Barasano mark gender and animacy in the third person; curiously, the inanimate marker is also used for speech act participants, i.e. first or second person, singular or plural (Jones & Jones 1991:73-4).
Tagalog (Philippine Austronesian) has an internally unanalysable independent personal pronoun expressing the meaning 'first person singular non-topic acting on second person singular topic' (Raritätenkabinett, on-line; contribution attributed to David Gil, LINGTYP List, 12 Sep 2000). Similar interaction effects - or anomalies in combinations of first and second person - have been found in three distinct geographical zones: Australia, Papua New Guinea, and North/Central America (Heath 1991, 1998; Evans, Jung & Brown 2004). They are found in languages in which subject and object are represented by contiguous affixes, in the parts of the paradigm which represent the collision of speech act participants (e.g. I > you, you > me). Instead of the segmentable forms found in most parts of the paradigm, these values are represented by a variety of non-iconic devices, such as the omission of one argument, the coercion of other person forms, or the use of portmanteaux. Evans, Jung & Brown (2004) have labelled this phenomenon 'Heath's problem'.
Kornfilt (2007) describes an interesting phenomenon in Sakha (Yakut), where in relative clauses the agreement marker expressing person and number values of the subject in the modifier clause is attached to the head of the relative clause, not to the predicate. Examples (from Nadezhda Vinokurova's & Jaklin Kornfilt's field notes 2001), with Kornfilt's annotation, include:
The fourth person in Eskimo. It has been claimed that there is 'fourth person' in Eskimo, however, it is better analysed as a long-distance reflexive.
Is there fourth person in Estonian? Grammars of Estonian frequently list 'fourth person' inflection in this language. This is the impersonal marker (of a morphological impersonal) which is better analysed as a marker of diathesis/voice.
Does Archi have the grammatical feature 'person'? Archi, a Daghestanian (or North East Caucasian) language traditionally assigned to the Lezgian group, has no unique forms for agreement in person, and the standard descriptions of this language do not involve the feature person (Kibrik et al. 1977; Kibrik 1977). However, the agreement patterns in Archi may be interpreted in favour of the presence of this feature, despite the absence of any phonologically distinct forms realising it (Chumakina, Kibort & Corbett 2006).
Archi distinguishes four genders and two numbers. The following table lists affixal agreement forms marking verbs in Archi:
The personal pronouns zon 'I' and un 'you (singular)' take gender agreements corresponding to the gender of the speaker or addressee: male humans trigger gender I agreement, female humans - gender II agreement, and imaginary locutors of genders III and IV (e.g. a speaking cow and a speaking goat kid) trigger gender III and IV agreements, respectively.
If Archi has no person feature, we should expect the same pattern of agreement, based on gender, to occur with personal pronouns in the plural. Indeed, this is what happens with the personal pronoun teb 'they'. It takes gender I/II agreement (the prefix b-) when the referents are human, and gender III/IV agreement (zero marking) when the referents are non-human. However, unexpectedly, the personal pronouns nen 'we' and žwen 'you (plural)' referring to humans do not take the gender-based I/II agreement marker (b-). Instead, they trigger zero marking, which we gloss as III/IV.PL as in the table above:
One possibility to account for this unexpected agreement is to say that the two pronouns nen 'we' and žwen 'you (plural)' are unusual lexical items. This approach was adopted in the standard grammar of Archi (Kibrik et al. 1977); apart from the four clear-cut genders (I-IV), four more genders (V-VIII) were proposed to account for different patterns of agreement triggered by some pronouns and a very small number of nouns involving human referents. The table below lists all the genders proposed by Kibrik et al. (1977) and Kibrik (1977). Furthermore, it additionally includes two more genders that have to be proposed to account for first and second person pronouns referring to non-humans:
Thus, an analysis of Archi without a person feature requires a complication of the gender system: when the cases of zon and un are taken into account, we have unique lexical items that belong to all four genders and require us to add four more genders to the system. Such an interpretation of personal pronouns seems to be unduly complicated and also counterintuitive.
Another argument against describing Archi verb agreement purely in terms of gender and number comes from the behaviour of conjoined phrases. For most combinations of conjuncts, the significant factor seems to be the presence of at least one conjunct denoting a rational. Thus, conjuncts headed by nouns in genders I and II trigger the I/II plural form of the verb, and when there are no conjuncts denoting rationals, III/IV plural agreement is found. However, when one of the conjuncts is a pronoun (from Kibrik's gender V or VI), the hypothesised resolution rule based on semantics does not produce the expected result. Instead of the predicted gender I/II marking (b-), the conjoined phrases trigger zero marking, the form equivalent to gender III/IV plural agreement. This is summarised below:
Kibrik's solution to the agreement pattern in coordinate constructions in Archi has been to group the proposed eight genders into ranks, with rank 1 comprising genders V and VI; rank 2 - genders I, II, VII and VIII; and rank 3 - genders III and IV. He then suggested a resolution rule, based on the system of eight genders and their ranks, according to which the target verb and auxiliary will agree with the gender of the conjunct belonging to the numerically lowest rank (rank 1 < rank 2 < rank 3). Thus, Kibrik's analysis of agreement patterns in Archi can be summarised as follows:
The rule accounts for all the examples discussed above, but it is typologically an odd resolution system. First, it is 'two-level', with genders and ranks of genders. Second, the reference to genders V and VI is essentially an indirect way of referring to personal pronouns, making them a kind of exceptional category within the gender system.
We should indeed base our gender resolution rules for Archi not on gender classes, but on a general rule formulated in purely semantic terms (i.e. if there is at least one conjunct denoting a rational or rationals, gender I/II agreement will be used; otherwise gender III/IV will be used). This much is compatible with what we know about systems of resolution rules, but it does not account for the agreement with conjoined pronouns.
Therefore, rather than treating the personal pronouns as each being an exception in terms of gender, we prefer to accept a person feature. If we recognise that there may be person agreement in Archi despite the lack of unique phonological realisation, both the behaviour of the two plural pronouns ('we' and 'you') and the resolution rules in general become simpler and cease to be typologically odd. Formulated in this way, the gender resolution rules are fairly usual, and the person resolution rules required (persons 1 and 2 > person 3) are standard, except for the interesting point that there is no distinction here between persons 1 and 2:
Thus, we analyse Archi as having agreement in person, with the following paradigm:
What is the person value of the pronoun one in English? Lexically, it has third person value. When it is used with first person meaning (for example in Queen's speech), it is still formally third person (One is... / *One am...), but - by convention, in certain circumstances such usage can be interpreted as referring to the speaker. This is analogous to other situations where - by convention - the interpretation of the person value may be different to the value normally expected of the lexical item. Examples include: Mummy is going to give you some cake, or Give it to Mummy, or Would Your Honour like to see the evidence now?, etc. Still, in all these examples, the relevant items are formally third person (assuming that nouns are third person by default), even though they refer to various participants in the speech act.
This can be seen as another instance of the distinction between syntactic/formal versus semantic agreement. In the English examples above, the nominals in question trigger syntactic agreement in person despite the meaning that may be (perhaps exceptionally) selected for them. But in analogous situations in some other languages, semantic agreement in person may be triggered in the predicate. For example, in Tamil we find sentences which are equivalent to English Mummy am going to give you..., and Anna am speaking when answering the phone. Corbett (2006:161) refers to instances of the latter as 'meaning-meaning' mismatches, because '[w]e have a mismatch between a name and its pragmatic use, normally treated in opposite ways in Tamil and English.'
Familiar polite plurals used when addressing a singular interlocutor could probably also be considered conventionalised examples of syntactic/formal agreement in person. However, we know that there are languages in which polite plural pronouns are genuine hybrids - i.e. they may trigger both plural (i.e. syntactic/formal) and singular (semantic) agreement in the same clause. Corbett (2006) cites examples from Czech and Bulgarian, where we find both plural and singular agreement in the verb phrase simultaneously (the following Bulgarian example is from Corbett 2006:231):
The problem of pro-drop. The controllers of agreement in person are linguistic elements that express syntactic arguments - these are typically nouns or pronouns, but may also be pronominal affixes. For detailed discussion of pronominal affixes, and the diagnostics for distinguishing pronominal affixes from agreement markers, see Corbett (2003), also summarised in Corbett (2006:99-112). Note that, on this account, pro-drop phenomena (where, in terms of syntax, it may be arguable whether a given element represents an argument or does not fill the argument slot) are not treated uniformly, but need to be analysed on a language-by-language basis, according to the criteria for identifying agreement in person.
How to cite this entry: