Common ground

This concept is in line with the general discussion of the inescapable perspectivation involved in linguistic construal and therefore has strong overlaps with Langacker’s construal operation of perspective. It also provides a link between linguistic construal and the notion of common ground to be discussed in the next section. Finally, the concept of constitution/gestalt refers to “the conceptualization of the very structure of the entities in a scene” (ibid.:63) and is linked to Gestalt psychology and phenomenology. This concept also lacks a straightforward counterpart in Langacker’s classification.

The category perspective/situatedness in Croft/Cruse’s model of linguistic construal operations contains the subcategory deixis, in which the authors introduce the notion of

epistemic perspective. This perspective situates the speaker and the hearer in a given communicative context with reference to “the shared knowledge, belief and attitudes of the interlocutors“ (ibid.:60). Croft/Cruse link this notion of epistemic perspective to the concept of common ground, which is widely used in CL to model the shared knowledge underlying communication within a given discourse community (e.g. Taylor 2002:346; Langacker 2008:466). Common ground can thus be directly linked to the classification of scientific and technical texts proposed in 2.7 since it provides a way of making statements about the different communicative configurations and the respective knowledge requirements in scientific and technical discourse from a cognitive linguistic perspective.

28 _{This is also consonant with the embodied realist rejection of a God’s eye perspective on the world (see}

The common ground concept was originally introduced in theoretical discourse by Stalnaker (2002:151, see also Clark 1996:93) but the major theoretical contribution to the concept is generally attributed to Clark (1996). Clark (ibid.:93) defines the common ground between two people as “the sum of their mutual, common or joint knowledge, beliefs and suppositions”. He further distinguishes between three representations of common ground, which are CG-shared, CG-reflexive and CG iterated (ibid.:94-95) and argues for CG-shared as the psychologically most plausible and most fundamental concept to be theoretically elaborated further.29

p [a certain piece of knowledge or information] is common ground for members of community C [e.g. a speaker and a hearer] if and only if:

The concept of CG-shared assumes a shared basis between two or more interlocutors “for the piece of common ground that some proposition

p holds“ (ibid.:94). The concept of CG-shared is formally represented as follows (ibid.):

1. every member of C has information that basis b holds;

2. b indicates to every member of C that every member of C has information that b holds; 3. b indicates to members of C that p.

This very abstract description of common ground becomes clearer if it is applied to a real- life example. Suppose that p refers to the location of the piston in a petrol engine, and the community C includes two engineers who discuss sulphur deposits on pistons. In order for the two engineers to assume that the location of the piston is common ground between them, they will look for a certain shared basis b that will justify this assumption. This search for a shared basis for an assumed piece of common ground is what Clark (ibid.:96) calls the “principle of justification”:

In practice, people take a proposition to be common ground in a community only when they believe they have a proper shared basis for the proposition in that community.

In general, there are various potential shared bases for a piece of common ground, and these will normally differ in how strongly they justify the relevant piece of common ground. This in turn is what Clark (ibid.:98) calls “quality of evidence“, which can be used

29 _{The notion of CG-iterated actually represents a prior representation of common ground that had to be}

discarded because of its cognitive implausibility (Sperber/Wilson 2_{1995:16 ff.; Clark 1996:96). The}

representation would look like this: A knows that B knows X. B knows that A knows that B knows X. A knows that B knows that A knows that B knows X, and so on ad infinitum.

to “rank” potential shared bases according to their strength of justification. A high-quality piece of evidence may be the physical co-presence of the two engineers working on a disassembled engine in which the piston location is clearly visible. The formal representation of CG-shared30

1. every member of C [both engineers] has the information that b [physical co-presence in the vicinity of a disassembled engine, piston location clearly visible] holds;

may thus look like this:

2. b [physical co-presence] indicates to every member of C [both engineers] that every member of C has information that b holds;

3. b indicates to members of C that p [location of the piston in a petrol engine].

Therefore, by making reference to a shared basis and ranking this basis according to its quality of evidence, we can assume, in communication, that a given piece of information known to us will also be known to our interlocutor(s) and is thus common ground between us. Linking the common ground concept to the idea of theory of mind illustrated in 4.3, we could say that by virtue of such high-quality shared bases, we can attribute very specific mental states to our interlocutor(s), for example the mental state of knowing a piece of information that is known to us as well.31

After this formal elaboration of his common ground concept, Clark goes on to distinguish two types of common ground, namely communal common ground and personal common ground (ibid.:100 ff.). What is important to the present discussion is primarily the notion of communal common ground. This type of common ground is closely linked to the notion of

cultural communities, which are “set[s] of people with a shared expertise that other

communities lack“ (ibid.:102). According to Clark (ibid.), it is constitutive of such a community that there is a “shared system of beliefs, practices, nomenclature, conventions, values, skills, and knowledge” about a certain set of phenomena. Examples of the bases of shared expertise that binds a cultural community together are nationality, residence, education, occupation, employment, etc. (ibid.:103). Applied to the above example of the

30 _{Note how the cognitively implausibly process of iterated knowledge attribution (CG-iterated) is avoided in}

CG-shared by making reference to an external shared basis that can be ranked according to its quality of evidence for the existence of a given piece of common ground.

31 _{Clark himself (1996:111) seems to establish an implicit connection between his common ground concept}

and the theory of mind when he claims that “we [...] have an intuitive feeling about what others know, which we might call feeling of others knowing [...]”.

two engineers, the fact that both of them had a similar university education or that they are employed by the same company and deal with petrol engines on a regular basis (and are therefore members of a common cultural community) could serve as further bases for the assumption that the piston location is a piece of common ground between them (which would probably need to be invoked if they were not physically co-present32

The common ground concept can thus be used to model the shared knowledge of a specific discourse community and, therefore, provides a link between the conceptual and the social dimensions of knowledge. Furthermore, as mentioned at the beginning of this section, it is not difficult to establish a connection between Clark’s notions of communal common ground, cultural community and shared expertise and the three-dimensional classification of scientific and technical texts proposed in 2.7. Both the different communicative configurations and the different degrees of technicality of scientific and technical texts proposed in this classification basically reflect different configurations of communal common ground between the authors and readers of such texts. In the present thesis, the common ground concept will have a two-fold application. Firstly, as described above, it will be used to model the shared knowledge underlying texts with different degrees of technicality. Secondly, it will be understood as the intersection of individual knowledge contexts (for example, the knowledge contexts of authors and readers of scientific and technical texts) and, as such, will serve as one dimension of context responsible for the relative saliency or centrality of encyclopaedic information in a given stretch of discourse (see 5.3). The actual organization of the knowledge underlying the discourse of specific discourse communities in the form of common ground between the discourse participants is modelled within cognitive semantics, to which we turn next.

in the vicinity of the piston).