- •2.4. Subjecthood in quoted imperative and hortative clauses.
- •1PlS / # go.Down-Hort
- •2.5. Transpersonal reflexives, imperatives, and hortatives in Russian
- •In spite of its (partial or full) suppression, the 1Pl subject binds reflexives in hortatives as in indicatives. In (37), sebe is a case form of sebja.
- •2.6. Addressees and allocutives.
- •1SgSubj now go.Out-Perf.SgSubj
- •References cited
In spite of its (partial or full) suppression, the 1Pl subject binds reflexives in hortatives as in indicatives. In (37), sebe is a case form of sebja.
(37) dava-j po-govori-m [o sebe]
give-Imprt Pfv-speak-1Pl [about Refl]
‘Let’s talk about ourselves!’
To summarize, Russian transpersonal reflexives are variably sensitive to agency. They are bound by hortative subjects as in Dogon and English. They are also bound by imperative subjects, as in English and most languages, but unlike Dogon. Russian hortatives are expressed by combining an imperative (with second person subject) with either a conjugated verb (with 1Pl subject) or an infinitive. Russian hortatives reflect the double nature of hortatives (as in Dogon and English), but because of their two-part form they do not challenge the validity of “subject.”
2.6. Addressees and allocutives.
The closest structural parallels we know to Dogon imperative addressee-marking of are the family of ethical (or personal) datives in general, and Basque allocutives in particular. Ethical datives of the Shakespearean and southern American English dialectal types differ from Dogon addressee marking both in limiting the dative to non-arguments of the core proposition and in allowing datives of any pronominal person, with first singular being especially frequent, as in Shakespeare’s He thrusts me himself into … (TG IV.iv). Basque allocutives similarly limit the dative to non-arguments, but like Dogon addressees and unlike ethical datives they are confined to second person categories. They occur in main clauses that are not already furnished with a second person pronominal argument (‘Patxi saw the donkey yesterday’ gets an allocutive, but ‘Patxi saw you yesterday’ does not). The combination of second person arguments and (in their absence) allocutives continuously indexes the socially relevant features of the interlocutor(s), viz., male/female, familiar/formal, singular/plural, and diminutive, the details differing from one dialect to another (Lafon 1959, Rebuschi 1981, Alcázar & Salterelli 2014). Allocutives are cut from the same sociolinguistic cloth as the T/V second-person pronoun oppositions in many European languages (Brown & Gilman 1960, Friedrich 1972), but unlike these pronouns they cannot be avoided.
Dogon addressee-number marking, limited to deontics, has no special sociolinguistic or discourse functions of this type, and comparisons to allocutives and especially to ethical datives should not be pushed too far. In imperatives, marking addressee number has the same communicative function as marking imperative subject number in any of the world’s 6000 other languages, e.g. French viens! versus venez! ‘come!’.
In hortatives, on the other hand, the Dogon distinction between actor/subject on the one hand and addressee on the other does make sense pragmatically. This is because the speaker hopes to be part of the soon-to-be-acting group, but the deontic illocutionary force is aimed squarely at coaxing the addressee(s) into joining up. The source of the modern English hortative (unreduced Let us VP!, surviving in archaisms such as the minister’s Let us pray!), did make provision for its bipartite nature by combining imperative let! with a first (inclusive) plural object denoting the actor group, defying the usual discomfort with sloppy coindexation of subject and object. However, the construction is now structurally opaque, as shown not only by the obligatory contraction to Let’s VP but also by the assortment of negative counterparts (Let’s not go!, Let’s don’t go!, ?Don’t let’s go!). Hortatives in many languages lack even a trace of imperative structure (vamos, allons-y, gehen wir).
3. Low-referentiality pseudo-subjects
Dogon languages also have another type of NP that looks at first like a normal subject but that fails some subjecthood tests. These are low-referentiality subject-like NPs, here called pseudo-subjects (we prefer this to the vague and over-used “impersonal subjects”), that occur in certain fixed subject-verb collocations. Some of the collocations denote ambient conditions such as time of day, season, or weather (§3.1). Others include a body-part term and denote emotional or somatic state (§3.2). In some of the latter, a true subject co-occurs with the pseudo-subject.
Intuitively, these pseudo-subjects are intermediate between impersonal it in It’s raining or It’s cold and fully referential subject NPs as in The children have gone to school. Even for the former, linguists have disagreed as to whether it is a semantically empty slot-filling expletive (dummy) required by the syntax, as in traditional grammar and some versions of generative syntax, or whether it has skeletal referential content. The latter viewpoint, argued most strongly by Bolinger (1977) and Langacker (2011) but also found in some generative treatments, is supported by conjunction reduction as in it rained and (it) thundered, and by the control complement in itx’s trying [x to rain].
3.1. Pseudo-subjects in temporal/meteorological collocations.
The Nanga examples in (38), denoting time-of-day, season-of-year, and meteorological states, have analogues in several other Dogon languages. They are classified as R-impersonals (i.e. with low referentiality) in the typology of Malchukov & Siewierska (2011b:4).14
(38) Temporal/meteorological collocations (Nanga)
noun verb gloss of collocation
a. time-of-day
kɔ̀rⁿɔ́ŋɔ̀yⁿ kárⁿí- ‘day break’ (lit. “morning happen”)
ùsí dɛ̀rⁿɛ́- ‘night fall’ (lit. “sun have spent mid-day”)
b. time-of-year
yàrí dɔ̌:- ‘cloudy weather arrive (onset of rains)’
yàrí gǒ:- ‘cloudy weather go away (end of rains)’
c. meteorological
bòndí wɔ̌:- ‘(it) rain’ (noun bòndí ‘rain’)
In ‘day break’ (38a) and ‘(it) rain’ (38c), the noun carries most of the lexical information. kárⁿí- is a semantically light verb (transitive ‘do’, or as here intransitive ‘be done, happen’), and wɔ̌:- is attested only in this collocation shown (we gloss it ‘rain.fall’ in interlinears). ‘Night fall’ (38a) is a more interesting expression, since it literally denotes the completion of the preceding time-of-day period (mid-day).
The two paired combinations with noun yàrí in (38b) denote respectively the onset and end of the buildup of rainclouds that characterize the rainy season, respectively around May-June and September-October (the rains peak in late July and August). Elsewhere, uncompounded yàrí means ‘sky’, but in these collocations it evokes compounds like yàr-kùmùrⁿá ‘cloud(s)’ and is best glossed as ‘cloudy/rainy weather’.
In Nanga, there is mixed evidence that the nouns in (38) are not true subjects syntactically. Fully referential NPs denoting specific individuals or sets normally precede temporal adverbs (‘now’, ‘yesterday’), as in (39a), while low-referentiality NPs with nonspecific meteorological or temporal reference usually follow (39b-c). However, the distinction is not categorical and alternative orders are possible in both cases.
(39) a. á:mádù níŋèyⁿ [ìsè gó] bírɛ́-ŋ̀
Amadou now [village in] work-Impf.3SgSubj
‘Amadou is working in the village now.’ (Nanga)
b. níŋèyⁿ bòndí [ìsè gó] wɔ̌:-ŋ̀
now rain(n.) [village in] rain.fall-Impf.3SgSubj
‘It’s raining in the village now.’ (Nanga)
c. yéŋìrⁿì: [ìsè gó] bòndí wɔ̌:-só-
yesterday [village in] rain(n.) rain.fall-Perf-3SgS
‘It rained in the village yesterday.’ (Nanga)
The situation is more sharply delineated in TgK, which (unlike Nanga) has a well-defined clause-initial position for true subjects (pronominal as well as nonpronominal). In TgK, true subjects precede setting-created temporal adverbs like ‘yesterday’, while pseudo-subject nouns in temporal/meteorological collocations follow such adverbs. Informant data and judgements were robust on this point for TgK. So there is hard evidence of a syntactic distinction between true and pseudo-subjects in TgK.
For example, both the pronominal subject in (40a) and the nonpronominal subject NP in (40b) consistently precede temporal adverbs like nɛ́:-wⁿɔ́ⁿ ‘now’. By contrast, the pseudo-subject pàrá ‘cloudy weather’ (semantically comparable but noncognate to Nanga yàrí) follows such adverbs (40c).15
(40) a. íⁿ nɛ́:-wⁿɔ́ⁿ gò-è
