8.2. Formal properties and main problems

The main characteristics of final detachments have been described as follows:

they occur mostly in oral/informal language; they have a special prosodical pattern (flat intonation). From a formal point of view, they contain a resumptive


pronoun or a clitic in the main clause; the detached lexical element follows the main clause and carries the casual marking of the element in question (in languages where this is possible). Different functions have been associated with these constituents: clarification, disambiguation, repair, recall of the main Theme etc. On the one hand, there are approaches that consider that this constituent is used in discourse above all for specifying the reference or correcting an error of formulation, while on the other is the viewpoint that they are programmed as such in the ongoing discourse, without any idea of repair or adjusting the reference.

Estonian scientific grammar (Erelt et al. 1993) treats these constructions as

‘specifying tails’, defining them as elements that add something to the clause.


Ta on tore poiss, see sinu vend.

‘He is a nice guy, this brother of yours.’

They are considered as not falling within the scope of the syntax, because there is no grammatical link between the two elements of the sentence. They have not been described in Estonian at the discourse level.

According to K. Lambrecht (Lambrecht 2001: 1068), final detachments (‘Antitopic’ in his terminology) are characterized by the following features:

from a syntactic point of view, final detachments are more closely connected with the predicate-argument structure of the clause than initial detachments. The main phrase containing the pronominal can be, in the case of initial detach-ments, extended by different types of clauses (complement, relative clause, etc), so that the nominal element and the pronominal are not adjacent; this is not possible in the case of final detachments.

Case marking in final detachments is more restrictive than in initial detach-ments: the nominal element bears the same case marking as the cataphoric constituent in the main clause.

There is also another constraint at the reference level: the detached con-stituent must be coreferential with the pronominal element, whereas in the case of initial detachments this criterion is less strict.

M. M. J. Fernandez-Vest argues that final detachment allows for identi-fication after the Rheme, considering that the detached element is not separated by a pause or hesitation and occurs typically in highly interactional contexts.

According to this author, final detachment seems to be triggered by a cognitive automatism, whereas the occurrence of initial detachment is mainly driven by the underlying principle of discourse activity that the first element is the one that one will be talking about (Fernandez-Vest 2006: 190).

She also underlines the fact that the final detachment does not stem only from the thematic part of the utterance and it is not only a “postponed Theme”:

this argument is supported on the one hand by the occurrences where the Mneme serves as a basis for the upcoming Theme and on the other hand by the


fact that a Mneme can also refer to elements presented in the Rheme (ibidem: 190–191). But Fernandez-Vest also includes other constituents in this category that occur in, for example, enumerations with plain intonation, having little informational charge and referring to the initial hypertheme (ibidem: 188);

in that case, the binary strategy Rheme-Mneme is present.

The prosodical difference between initial and final detachments is explained by K. Lambrecht by their different syntactic connectedness: the initial detach-ments are syntactically independent, whereas final detachdetach-ments are syntactically connected with the preceding elements and do therefore not bear a clear accent (Lambrecht 2001: 1071). The questions related to the prosody will be addressed in the next section.

Final detachments are somewhat similar to some of the constructions that Discourse Analysis considers as a certain type of repair mechanism (after-thoughts). These two types of constructions have been analysed as different ones, for example by Ashby (1994) in French and Fretheim (1995) in Nor-wegian; according to K. Lambrecht (ibidem: 1076), who also refers to the arguments of Ziv (1994), afterthoughts constitute separate intonation units, they are thus accented and preceded by a pause – which is not the case for final detachments. Also expressions or particles that refer to the repair function in afterthoughts, should help to make the difference. This controversy will be discussed further, along with examples from spoken Estonian. Some researchers tend to think that one should not consider together two radically different constructions: final detachments (planned as such, as part of the speaker’s

‘grammar’) and repairs or afterthoughts (occasional formulation work). In addition it is admitted that there are many constructions that have some formal properties in common with typical final detachments, such as apostrophes, appositions, parentheticals, but which do not fill all criteria that allow them to be considered as final detachments (for discussion see Apothéloz & Grobet 2005: 98–102).

Different constituents can be indeed detached after a clause; in this sense, there are some categories that we will not investigate here, for example, infinite constructions etc., also vocatives are considered as falling outside of the scope of this study.

In Estonian, as in English, the stressed forms of pronouns do generally not occur alone in detached constituents, whereas in French it is a strongly gram-maticalized phenomenon, to the point that the construction with moi je in initial detachment is not considered as a detachment any more. In English there is the possibility of using the stressed pronoun I, as with the longer form of the pronoun, mina, in Estonian. In French, the stressed form can not be used without the clitic je.

In languages with relatively flexible word order like Estonian one can raise the question about the possible variation of word order when considering final detachments: in interrogative utterances in spontaneous oral language, the subject is often placed at the end of the utterance; there can be a resumptive


pronoun, so that the utterance can be considered as a final detachment, but the pronoun is not compulsory, in which case the utterance follows the quite regular pattern in oral language of placing the subject at the end. This tendency has been examined by L. Lindström (2002: 102), who has explored the conditions which favour the word order VS in spontaneous oral language. According to her, the following conditions favour the apparition of the VS order in Estonian, one of the most important criteria being the ‘agentivity’ (the scale going from first- and second-person pronouns to full NP referring to inanimate referents):

The utterance begins with a possessor adverb The subject phrase refers to a non-human The subject phrase is a full NP

The subject phrase conveys the new information in discourse

The next example fills most of the conditions in this regard: the possessor adverb tal (he-ADE), the subject is not human (operatsioon, ‘the operation’), the subject phrase is a full NP (see operatsioon) and in this conversation the reference to the operation intervenes for the first time (the marker aa at the beginning of the utterance indicates the introduction of a new item, or the fact that the speaker recalls an element), although the participants seem to be familiar with the event in general.


V1: {aa.} no okei siis. (0.5) .hh a ‘meie lähme üldiselt vist PRTCL okay then. we go.1pl generally probably

‘maale ära präegu nii=et noh [me] lähme ’Kasparit countryside.ALL away now so that PRTCL we go.1pl Kaspar.PART vaatama.


H: [mhmh]


H: aa kuidas tal ‘läks see operatsioon.

how he.ADE go.PST.3sg DEM operation

V1: ‘Okay then. We will go now in principle to the countryside, so that we will see Kaspar.

H: Uhuh.

H: How did it go, his operation?’


In this sequence the interrogative utterance does not contain a resumptive pronoun, but the use of see in the main clause is absolutely possible (kuidas tal see läks see operatsioon).


In oral language the VS word order corresponds to a general tendency of giving first the predication followed by the subject of the predication. How, then, should one make the distinction between final detachments and variation of word order in Estonian? In the present thesis the examples without the resumptive pronoun were in general not considered as detached constructions, as we do not yet have clear-cut criteria permitting making a distinction between the two, if it were justified to do so. In some cases, there is a clear pause before the detached element, but as this feature does not occur systematically in final detachments, we cannot establish a rule for defining all cases of detached con-structions. Moreover, the constructions without the resumptive pronouns should be examined on their own, as various constructions can be found under this category, taking into account also other formal properties, for example in inter-rogative/assertive utterances (prosody), which should be analysed separately.

Im Dokument MARRI AMON Initial and final detachments in spoken Estonian: a study in the framework of Information Structuring (Seite 146-150)