unit 3: lexical density analysis

LexGram | Stephan Hegglin | June 3rd, 2001

I have also just got to the Lexical Density Analysis in my TDA Module Unit 3. I was wondering why Bloor did not point at the difference between "to be" as an auxiliary (grammatical term) and "to be" as a lexical term as Sheehan has already explained very convincingly.

If we just solve the problem of categorizing grammatical terms with their affiliation to closed sets (word-classes with fixed numbers of words), without considering more subtle differences, it's not quite comprehensible why in the nit-picking approach a little and in place of are given so much attention.

Do you find the knit-picking approach convincing?

I'm looking forward to hearing more about this.

All the best

Stephan Hegglin

Re: 'Tis | Raymond Sheehan | June 4th, 2001

"It depends on what your definition of the word 'is' is." (President Bill Clinton)

See also the title of the sequel to "Angela's Ashes".

Raymond Sheehan

Re: 'Tis | Stephan Hegglin | June 4th, 2001

To Raymond: I have read "Tis" and not taken the hook. Thanks.

Stephan from sunny Switzerland

Re: Lex Gram | Tom Bloor | June 6th, 2001

Dear lexico-grammarians,

The debate about 'be' is not as clear-cut as either side seems to suggest. I would have thought that a reading of the persuasive case that you have made on both sides makes this clear. I don't know where, but somewhere in my materials I do discuss the issue of 'be' as grammatical/lexical and I say more or less what the lexical-'be' party have said but more tentatively: copula 'be' MAY be considered as lexical and the extremely rare instance of be = exist (I think therefore I am; God is!) must be lexical (This last point has to be right, and I leave it to one side). If anyone locates these comments of mine, I'd like to know where they are.

But many people -such as Nathan (on this list earlier)- would not see it that way. I think that Halliday's grammar indirectly suggests that he would agree with Nathan's view of both kinds of 'be' as grammatical rather than Stubbs's distinction (but Halliday doesn't talk about lexical density as far as I know). I can't explain this last point because it would be too confusing if you haven't done SFG in some detail; but for those of you who have done SFG, it's the fact that Halliday analyses all finite instances of 'be' as F rather than F/P, which suggests to me that he sees them as purely grammatical.

There is also evidence from stress and related phenomena (finite forms of "be" are usually unstressed and 'be' can be criticized (attached to the subject, eg "she's", "I'm") in the copula as well as the auxiliary function. Telegraphic speech can omit copula as well as auxiliary 'be'. I think that, for what it's worth, copula 'be' fits most of Stubbs criteria for grammatical items and not for lexical ones. Also there is arguably some cross-linguistic evidence to bare out the grammatical hypothesis, namely the fact that some languages can omit copula 'be' altogether. I forget which: Arabic? Some Slavonic languages? (Any informed polyglot out there?). Pidgins and creoles which tend to drop a lot of grammatical morphemes but not usually lexical items may drop copula as well as auxiliary 'be' e.g. English: He/She is bigger than you. Krio: I big pas u. French: Il/elle est plus gros(se) que toi. Seychellois: li gros pas u.

Incidentally, the argument that if a word has 'meaning' it must be lexical is not valid. Obviously, grammatical items must have some meaning or why they exist at all. Most (all?) grammatical items obviously have a meaning of some sort. Their presence or absence can greatly affect the content of a message, often at the crudest level,. But affixes have meaning too - and they are indisputably grammatical ("climb+ed" does not mean the same as "climb". A verb with n't added to it can mean something very different from the same verb without n't). I'm dubious about the extra-linguistic (sic) reference point, too. Don't most personal pronouns have extra-textual reference in the sense that they refer to some entity in the real or fictitious world?

Re Stephan's critique below (Hey, Stephan, this doesn't look like a barrel of Guinness to me!): One reason why I discussed 'a little' and 'in place of' is that in discussing this and similar data with students I had found that they were hotly debated. One reason for this is that they illustrate the important point of whether there is a case for breaking down lexical chunks into single words - questionable in grammatical analysis for most purposes but probably absolutely necessary for the lexical density approach to work at all. Hence, 'to give' is obviously a single grammatical constituent, but it is two words - one G and one L. That's what LD is all about. How many of the WORDS are G and how many L? (By the way, my unoriginal metaphor was "nit-picking", ie picking out the eggs of head-lice, but I suppose picking at knitting makes at least as much sense and is more original).

I am grateful to Barry Walford for succinctly stating the intention behind the lexical density analysis activity in FND, I quote:

"My impression of the FND exercise is indeed that we are expected to recognize the issues (and that there are issues) and justify our response to them, not that we discover some "right" answer: in many instances, there clearly isn't one, only a position to take."

That's exactly the point. The intention is to get people arguing about lexico-grammatical distinctions -thinking lexico-grammatically- but in the context of a problem that most people have not met before, and it seems that we have succeeded. If in the end, you decide that lexical density is not for you, no one at Aston will mind. Even if you decide it is garbage, no one will object if you can make a convincing case. Evidence is the name of the game. Barry's words should be inscribed on your hearts. The possible disadvantage of using the list is that you accept someone else's analysis as the correct one and fail to raise the queries that you see as a hindrance to completing the task but in fact constitute a crucial part of the task: the identification of problems. If I or someone else tells you what the correct decision is, the task loses most of its value. However, the advantages of discussion outweigh that danger.

Obviously, you have to "take a position" in deciding how to make your analysis, but your doubts can be expressed in the commentary. There is no reason why you should commit to a belief in LD or a rejection of it. I'm pretty ambivalent myself.

Keep debating,

Tom Bloor

Re: LexGram | Stephan Hegglin | June 6th, 2001

Thomas Bloor was wondering where the comments on be = exist can be found. The comments in "Further Study" in: Bloor and Bloor, The Functional Analysis of English (pp 58-60) proved to be very helpful. I don't know if these are the comments you were looking for.