מִפֹּה לְפֹה

Discussion questions

• This article deals with cislocative (‘venitive’) markers (‘hither’) changing into person markers. Are there cases of translocative (‘andative’) markers (‘thither’) changing into person markers? Depending on the deixis of the marker (‘near the hearer’, ‘away from the speaker’, ‘near somebody else’, etc.) I would presume there might be cases in which translocative markers can change into either second or third person markers.

• Given that in many languages there is connection between the personal deixis and the deixis of demonstrative pronouns and cis-/translocative markers (e.g. English here and this are proximal, 1-centric; there and that are distal. Japanese こ- ko-, そ- so-, あ- a- system of correlatives, loosely corresponding with 1st ‘near the speaker’, 2nd ‘near the hearer’ and 3rd ‘distant from both’ person deixis), how common is it for markers to cross categories?

• A theoretic, meta-linguistic question following the previous one: should we speak of a unified super-system of personal deixis in languages in which demonstratives, markers of location, etc. are aligned along personal deixis (like the Japanese ko-, so-, a-)?

Critical review

Summary

This article explores one source of personal pronouns which is dealt with less frequently, relatively, in the linguistic literature: cislocatives (elements meaning ‘hither, (to) here’). It does this by describing the diachronic situation in three unrelated American language groups (Sahaptian, Shasta and Iroquoian), and discussing whether in each of these case studies a cislocative has truly evolved into a referential personal pronoun, as opposed to an implicit reading (cf. English give it here, implying ‘give it to me’ in this sentence while here can by no means described as a first person pronoun). Each of the three cases shows a different initial state and a different outcome:

• Sahaptian. These languages have a zero marker (called a ‘paradigmatic gap’ in the article) for indicating the interlocutors (SAP, first and second persons) in the verbal morphology (although they can be indicated by enclitics whose optionality differs among the languages in the group). Here the cislocative suffix, which can be traced back to Proto-Sahaptian *-ɨm, was possibly re-interpreted as a first person O or R marker.

• Shasta. Here the system of morphological person markers is richer, yet object pronouns have not been morphologized. The data suggest a set of cislocative suffixes was evolving into markers of first or second person O (or R).

• Iroquoian. This language family has a very rich system of pronominal affixes (50–60 in each language!) already established. This case shows how cislocatives can ultimately result in fully referential pronominal affixes integrated into such a system, indicating 2→1 (second person acting on first person) action. Several kinds of evidence are given to support this claim.

Critical assessment

In context of our course, this article is given as an example for one source of person markers (others being antipassives in Bickel and Gaenszle (2015) and additional sources in the optional reading). A typological overview of sources for personal pronouns is given in Heine and Song (2011), who deal with spatial deixis markers as sources for first and second persons in §2.2.4 and §2.3.1.1 (note 29 is especially relevant).

I found the article in question clear, convincing, interesting and enjoyable to read. It deals with situations which are not clear-cut by discussion and openly presenting the available evidence (including presenting competing analyzes, such as Rude’s (1990) analysis of the Nez Perce marker as an inverse maker, indicating an unusual (‘inverse’) direction in the flow of the action). The cases of Nez Perce and Shasta - even though it is not decisive in them whether the cislocative has in fact developed into a fully referential personal pronoun — can suggest how the described language stage could be set for the development of such pronouns. In the clearer case of Iroquoian, several kinds of evidence are given. The first one — speaker intuition — is quite problematic on its own in language description, but since it is one of several it can be regarded providing additional, ‘extralinguistic’ support.

The key concept is reinterpreting or reanalysis, which is often central in category shift. The shift here is not between two unrelated categories but withing two which are deictic. There are languages that even align the spatial deixis along with personal deixis (such as the Japanese ko-, so-, a- system). A broader, typological, cross-linguistic survey is needed. This survey might touch upon translocative markers, if there are cases in which they develop into person markers.

The place of politeness and sociolinguistics in the development of person markers, discussed in the Iroquoian case, seem to be central in many languages spoken by diverse language communities around the world. If we accept Bickel and Gaenszle’s analysis (which I’m reluctant to do; the route they present seems to me quite far-fetched and not supported well enough), that would be another case of the interaction between politeness and the use of person markers.