Appositions and Construction Grammar
So far, only a few types of appositions have been analyzed. I restricted the analysis to wide appositions of the type NP + NP and contrasted monological and dialogical passages of spoken interaction. Even then, to get a fuller picture it would be necessary first to extend the analysis to a much larger range of communicative genres (instructions, lectures, counseling sessions, consultations etc.) as well as to include both monologically oriented (books, magazines, legal texts, scientific texts etc.) and interactionally oriented (chat, e-mail, SMS etc.) written language.
In spite of the meager database of only 23 cases, though, some tentative hypotheses about how a Construction Grammar approach to appositions might look like might be ventured:
(i) First, prototypical wide appositions - excluding repair-like cases as presented below in (iii) - are centered round a noun phrase referring to a person (seven out of eight instances). It would be plausible, then, to assume at least one construction called “wide apposition: person + information about person”. The formal entry of this construction would be that there is a core noun phrase which is filled by a proper name or a word/phrase referring to a human being (his wife, my friend, their boss etc.) and an appositive noun phrase that follows the core and gives additional information about the status, function or profession of that person. The functional entry would be that this type of apposition is used to provide recipients with the necessary pieces of information to identify the person named.
The sequential entry - an additional type of information that would be necessary to add from the perspective of Interactional Linguistics (see Imo 2007a: 40) - would be that this construction is used preferably in monological situations, where a pre-planning of utterances is possible and where the apposition is a kind of “pre-emptive” attempt to prevent recipients’ lack of understanding. On a prosodic level, this type of apposition is usually realized within a separate intonation phrase, which reflects the ambiguous status of the apposition. On a syntactic level, it is integrated into an utterance and forms a single constituent together with the core noun phrase it refers to while on a prosodic level the information given in the apposition is marked off as an aside or as background information.
(ii) Second, there might be another type of wide apposition that could be called “wide apposition: thing/concept + information about thing/concept”. This assumption is rather a speculative one at the moment, though, because only one instance of this type was found in the data I searched and this occurred within a TV newscast for kids. The basic function is the same as in the first type of wide apposition. It helps prevent a possible lack of understanding and needs a monological background. An open question would be whether this type of apposition is usually realized with a discourse marker such as also (i.e.; ‘that is’), und zwar (‘namely’), das heift (‘meaning/that means/that is’) etc. or not. Another open question would be whether appositions referring to a core noun phrase denoting a thing/object are more widespread in other communicative genres, such as instructions, lectures etc.
(iii) Third, there is a type of “very wide” or “peripheral” apposition that occurs in interactional settings. It may be called “very wide apposition” because, on a formal level, other words or phrases may occur between the core noun phrase and the appositive noun phrase and, on a functional level, these appositions do not provide background information but rather re-focus, paraphrase or even, at the end of the spectrum, repair the previous utterance (see also Meyer 1992: 74). They occur in interactional settings because they are a result of the ongoing, temporal process of speaking. This temporal process can lead to the choice of wrong or at least less fitting words, which then have to be corrected or “fine-tuned”. Another reason is that the absence of recipient reaction (no taking over of a turn) may inform the speaker that further information is needed for the recipient to understand his utterance.
Because of this closeness of appositions and repairs, persons as well as things/concepts can constitute the core noun phrase of “very wide/peripheral” appositions, because anything may be a candidate for repair. At the far end of the spectrum, the apposition construction merges with that of repair. How closely “very wide/peripheral” appositions and repairs are related may be illustrated by lines 290/291 of example 9:
290 ^ durch (.) durch diese (.) durch diese Atemgerausche,
by (.) by those (.) by those breathing sounds’
291 ^ die SAUGgerausche,
the sucking sounds’
Preceding the apposition / repair “diese Atemgerausche / those breathing sounds” and “die SAUGgerausche / the sucking sounds” are two further self-repairs, namely “durch / by” (line 290), which gets replaced by “durch diese / by those” and then “durch diese / by those”, which in turn gets repaired by “durch diese Atemgerausche / by those breathings sounds”. These two self-repairs already hint at an ongoing word search by the host, giving further indications that “those breathing sounds / the sucking sounds” is more of a repair than an apposition proper. Self-repair may be conceptualized as a construction, too, which is defined by a process of breaking off during the production of an utterance and the replacement of an element produced before. Self-repairs are structurally very open. Prepositions can be replaced by prepositions, adverbs by adverbs, adjectives by adjectives, nouns by nouns but also parts of noun, adverb, prepositional or verb phrases by other parts (for a detailed analysis of how self-repairs work in German see Egbert 2009: 53-98). Only when a noun phrase is replaced by another noun phrase do amalgamations between the self-repair construction and the wide apposition construction occur. If only parts of a noun phrase are realized, the constructs move towards repair proper in the network of related constructions. If the second noun phrase does not plausibly repair the first one (i.e. does not replace it), the given construct moves towards a prototypical wide appositive construction.
(iv) Fourth, next to amalgamations of appositions and self-repairs there are also amalgamations of narrow and wide appositions, as in example 5 (“deine FREUNdin, die ANna / your girlfriend, the Anna” in line 165-166 and “deine FREUNdin die Anna / your girlfriend the Anna” in line 175), where stylistic (personalization strategies) and prosodic aspects combine to create a special case of a rather open formal and functional pattern whose constructional status is very unclear.
What this qualitative, empirical study has shown is that, on the one hand, Interactional Linguistics can indeed profit by taking up ideas and concepts of Construction Grammar. One problem of Interactional Linguistics is that analyses within this framework often remain focused on single phenomena which are described very fully but which are not embedded into a larger theory of syntax. Construction Grammar, on the other hand, forces one to think in terms of a coherent system of representation, and its concepts of the symbolic construction and the network of constructions deliver working hypotheses with which to achieve such a representation. The result of almost all previous attempts at combining Interactional Linguistics and Construction Grammar, though, has been that there is an enormous lack of knowledge about the actual properties of syntactic phenomena. So far, most linguists - and this holds true for Construction Grammarians, too - have relied on little or even no empirical data and an overhasty classification of syntactic phenomena, which means that every time a Construction Grammar analysis is attempted within the framework of Interactional Linguistics the lack of empirically supported knowledge about neighboring phenomena - which is necessary for the establishment of network-based descriptions - is felt severely. What this means is that much more empirically driven work about the actual use of syntactic patterns in spoken and written language is needed before a fully satisfying Construction Grammar representation of phenomena is really possible.
-  Meyer (1992: 98) discovered a striking inequality in the distribution of appositions across different genres within his corpora: “The genres of fiction and conversation contained the fewestinstances of appositions [...], the genres of learned writing and press writing the most [...]. Thisskewed distribution existed in the corpora because appositions are communicatively morenecessary in some genres than in others.”
-  Meyer (1992:132), too, states that the concept of “apposition” becomes fuzzy at the borders.In his data, he had quite a lot of apposition-like candidates that fulfilled only some of thecriteria for appositions proper so that he decided to differentiate between more “central” andmore “peripheral” appositions, according to how many of the six semantic and syntactic criteriathey fulfilled.
-  See Egbert (2009:101) for a description of the interactional status of such reformulations.