Assignment: Evaluate the explanatory power of Cognitive and Semiotic Grammar with regard to damn, so-called and fuckin' in this damn spoon, a so-called expert and those fuckin' chickens.

Preliminary note: References to Halliday 1994, Langacker 1991 and McGregor 1997 are indicated by the abbreviations IFG, CG2 and SG respectively.

The aim of the present essay is to assess the explanatory power of Cognitive (CG) and Semiotic Grammar (SG) with regard to the "attitudinal Epithets" (IFG:184) in the examples under (1). It will be argued that CG lacks the conceptual tools to incorporate this phenomenon in its analysis of nominals as presented in CG2 (Ch.2; Ch.3), whereas the framework offered in SG is flexible enough to deal with this issue in an enlightening way.

Now consider (1):

In terms of Langacker's definition of an instance - "an instance (but not a type) is thought of as having a particular location in the domain of instantiation" (CG2:57) - it is hard to image a domain of instantiation with, apart from a spatially located silver spoon, gold spoon, small spoon etc., also 'damn spoon'.

This leaves us with the function of grounding to be considered. At first sight, this seems a reasonable solution, if we consider this broad definition: grounding "pertains to the relation between the designatum and the speech-act participants" (CG2:51). Indeed, it seems correct to say that damn, so-called and fuckin' in (1) are uttered by the speaker in order to convey his attitude towards the spoon, the expert and the chickens respectively, to the hearer. However, apart from the fact that grounding normally presupposes instantiation – which is missing – I see at least two problems in granting them a grounding function: (i) the position of these elements disturbs the iconic layering of the functions as represented by the formula (G(Q(I(T)))) (cp. CG2:421) and (ii) the more technical definition of grounding seems to exclude them from the grounding function.

(i) Langacker notes that the four semantic functions he distinguishes "tend to be reflected iconically in a nominal's structure" (CG2:54). Indeed, the type specification is to be situated at the far right of a prototypical nominal, preceded by the instantiating, quantifying and, at the far left, grounding parts. If damn were to ground the instance this spoon, it would normally have to be in front of this: *damn this spoon (in the non-clausal reading). Even if we were to suppose that this and damn perform the grounding function together, the iconicity would still require damn to precede this, as this also quantifies and instantiates. Admittedly, Langacker (ibid.) allows for 'non-canonical' nominals such as nominals without a head noun or several functions incorporated in one word (the extreme of this is the proper name). He does not, however, mention the possibility of breaking through the ordering G>Q>I>T.

(ii) Langacker distinguishes two types of grounding predications: those which identify an instance with reference to a more inclusive class (reference mass R_T), and those which focus on degree of definiteness (i.e. is the instance presented by the speaker as known by the hearer) and take the ground (speaker and hearer) as their reference point (CG2:89;96). Of these two, damn, so-called and fuckin' can only be concerned with definiteness, as they do not pertain to quantity. In addition, they do not convey any information about definiteness, as appears from a comparison of (1a), this damn spoon, in which the hearer is clearly supposed to know (possibly see) the instance talked about (ii), to (1b), a so-called expert, in which a new entity is introduced in the discourse. Finally, if we turn to Langacker's technical definition of a grounding predication, it becomes very difficult indeed to attribute a grounding function to the elements under discussion:

In short, it appears from our discussion up to this point that Langacker's theory of four functions cannot accommodate such adjectives as damn, so-called and fuckin' in (1). The only function which might possibly be attributed to them, is that of grounding, but this possibility seems to be ruled out by his exact definition of this function. (One might of course choose to consider them an anomalous category of grounding predications.) Let us now turn to SG to see whether this approach can account for the fact that the words we are dealing with are not used referentially, but serve to express a speaker's attitude towards a discourse referent.

First, however, I would like to refer briefly to the only clear semantic label I have given the words under consideration so far, viz. attitudinal Epithets. In his discussion of the experiential structure of the NP, Halliday allows for a subdivision within the category of Epithets: they can either indicate (experientially, or, for our present purposes, referentially) "an objective property of the thing itself" (IFG:184), or (interpersonally) "the speaker's subjective attitude towards it" (ibid.). Interpersonal or attitudinal Epithets are not potentially defining, as is illustrated by (5a), in contrast with the experiential Epithet (iii) in (5b):

Some of these suggestions will return in our discussion of the SG account of these attitudinal modifiers. Indeed, SG would posit that the words printed in bold face in (1) entertain a conjugational relationship (the interpersonal semiotic). That there is at least something funny about these words, is already indicated by their different grammatical behaviour compared to regular adjectives maintaining a logical relationship of elaborating hypotactic dependency with an N (such as small in a small spoon). Some intriguing differences are listed under (6). (iv)

The type of conjugational or interpersonal modification involved in (1a) and (1c) is attitudinal or "speaker oriented, concerning the speaker's attitudes to the content being expressed" (McGregor 1998). In terms of the more syntactic, binary contrast between scoping and framing, I would argue that the attitudinal modifiers in (1a) and (1c) involve scoping. This implies that damn in (1a) and fuckin' in (1c) have attitudinal scope over this spoon and those chickens respectively. (7) offers a diagrammatic representation of (1a) indicating this scopal relationship (ATT=attitudinal).
 

(7)

 
(9)

First of all, it should be noted that swear words such as bloody, damn or fuckin' do not necessarily involve a negative attitude of the speaker. Consider (10), an imaginary conversation between pals:

The possibility of these swear words to occur within a word would certainly pose insurmountable problems to a CG analysis in terms of the four functions proposed by Langacker. SG can account for a phrase like those Aladamnbama farmers in the same way as for those damn Alabama farmers, or this damn spoon (cp. 7):

 
(12)

I would suggest, then, a kind of continuum of adjectives, ranging from ones that only serve as logical qualifiers, to ones that only serve as attitudinal modifiers. At the one extreme, you might find adjectives such as red or electric: there are objective criteria to determine whether or not something is red or electric. This is not to say that no doubt may arise as to whether something is, for instance, red or orange, but there are exact scientific criteria to distinguish the two colours (in terms of their values for red, green and blue). I would situate the swear words such as damn and fuckin' at the other extreme, because they only express the very 'private' attitude of one particular person at one particular occasion. In between these two extremes, adjectives such as lovely, ugly, stupid and the like would be situated. I do not mean to suggest that adjectives can be divided into three categories: the notion of a continuum implies that all sorts of intermediate stages are possible. Consider the adjective hot, for instance. Although heat is a measurable quality, what is hot for one speaker, may not be so for another, and there is no scientific definition of hot (only of heat). This suggestion of a continuum is represented in diagram (14) (vii). Incidentally, it would appear to be a relative universal of language to have the (objectifiable) logical qualifiers closer to the head noun than the attitudinal modifiers ("Subjective-Objective gradience", Hetzron 1978) – but as we have seen, the swear words in particular do not readily conform to word order rules.

 
(18)

Notes

(i) For instance, whereas fucker 'fool, idiot' does have a definition in the Oxford Advanced Learner's Dictionary (1989⁴), fucking only receives a note on usage: (used to add emphasis in expressions of anger, annoyance, etc). [Back to text]

(ii) Of course, this can exceptionally (viz. in colloquial speech) serve to introduce a new referent, as in There was this strange bloke, and he came up to me, and he said 'boooh'. It is, however, hard to imagine a context in which There was this damn spoon might feature, except perhaps in a twisted fairy tale. Even so, this usually refers anaphorically to an entity previously 'mentioned'. (It seems to me that this mentioning need not be verbal; one might imagine someone setting the table and dropping a spoon several times exclaiming "This damn spoon!". If any hearers are present, they already know about this spoon because they have seen and/or heard it.) [Back to text]

(iii) That is, in Halliday's view. In SG, long in the long train does not serve an experiential role (if it would, it would have to be Entity, but that role is clearly taken in by train). Instead, it is in a hypotactic dependency relation of elaboration vis-à-vis train. [Back to text]

(iv) This table is based on the suggestions made by Prof. McGregor in the seminar on SG, regarding the differences between the Epithets and Classifiers of Hallidayan grammar. This is not to suggest that damn in (1a) is a Classifier, of course. For one thing, there is not a limited, "mutually exclusive and exhaustive" (IFG:185) set of spoons in terms of how they are appreciated: you cannot say, for instance, that * a spoon is either damn, bloody or fuckin'. In addition, note that there are other types of adjectives that can only be used attributively, such as postdeterminers (usual, famous, etc.) and the 'anomalous' category of chequered, belated, knotty etc. (Davidse 1996:5.7-5.8) [Back to text]

(v) It is possible to paraphrase (1b) in a sentence of represented thought, and the putative heir in a sentence of represented speech, but paraphrases (8a) and (8b) seem more appropriate. [Back to text]

(vi) It is very difficult to come up with good criteria for these semantic types of conjugational relations, as Prof. McGregor readily concedes (pers.comm.). (8b) in fact suggests rhetorical modification (presupposition, truth value status) for the putative heir. Similarly, consider example (i):
                             (i) I spoke to your brother - I hope - on Saturday. (SG:250)
In SG (ibid.), the clause I hope is said to modify your brother attitudinally. I would be inclined to call this rhetorical, rather than attitudinal, modification: you "hope" that your presupposition that the person you spoke to was the brother of your current addressee, is right. [Back to text]

(vii) It seems to me that most or all classifying adjectives would appear at the extreme left. In addition, it appears to me that at least some postdeterminers or post-deictics (IFG:183) cannot be included in this continuum, as they seem to subcategorize without really qualifying or modifying. Consider, for example, certain problems, various ways, the usual excuse. Consider also the post-deictics referring to (dis)similarity: it is not an inherent quality of the train in the same train that it is the same. It can only be the same as some other train present in the discourse context. Indeed, SG would probably not posit the conjugational relation of elaborating hypotactic dependency, but a linking relation, viz. indirect reference (cp. SG:322). [Back to text]
 

Reference list