defining text
Defining text | Mary Lynn Hughes | May 16th, 2002
I'm looking at how definitions and clarifications of lexical items are signaled and effected within text, using three 'general' corpora, with maybe a look at the more restricted domain of the MSc list (if I can fit it in). At the moment, I'm trying to identify lexical items or lexical-syntactic patterns that commonly signal a definition or clarification and hope to identify some of the variety of ways definitions/clarifications are pragmatically effected (i.e. not just conventional definition formulae).
Here's a few examples from a newspaper corpus- data from a search on the keyword 'one' + 'of' (rhs) What I take to be the 'defining' text is underlined, with my interpretations below:
1) Differing interpretations of the legislation (a) is just one of the problems uncovered in a Law Society survey of family law solicitors relying on the Domestic Violence Act(b).
(a) 'one of the problems (uncovered in a Law Society survey?)' defined by 'differing...legislation' and (b) 'Law Society survey' defined by 'of family ... Act'.
2) With one of the most underdeveloped musical infrastructures in Europe (a) - in education, Northern Ireland is luxuriantly resourced by comparison (b) - this State has yet to define or declare a national music policy.
(a) Is the 'defined' simply 'this State'? what about 'yet to define ...policy'? (2) 'Northern Ireland' defined by 'luxuriantly...comparison'.
3) And if that wasn't enough, both Fegan and his friend and drug-dealing associate (c), Brendan "Bap" Campbell, were also involved in a running row with the Provisional IRA (b), regarded as one of the world's most efficient and deadly terrorist organisations (a).
(a) 'the Provisional IRA' is defined by 'one of the world's ... organisations' - also signaled by 'regarded as'. (b) 'Fegan' and 'Brendan "Bap" Campbell' - defined (or their role clarified) by 'also involved in a running row with the Provisional IRA', plus 'drug-dealing'? (c) 'Brendan "Bap" Campbell' - also defined by 'his friend and drug-dealing associate'?
4) Great players (a) remembered by one of the greats (b) [headline]. In a tribute to Gene Sarazen, who died last Thursday (a), Dermot Gilleece recalls these magic moments from an interview with him during a dinner in his honour at The K Club, in July 1993.
(a) 'Gene Sarazen' is defined/identified by 'died last Thursday' and 'great players (remembered)'. (b) 'Dermot Gilleece' is defined by 'one of the greats'. But what about 'these magic moments' - clarified by 'from an interview ... July 1993'?
5) Abruptly ending one of the shortest-lived governments in modern Dutch history, Prime Minister Wim Kok's cabinet resigned yesterday, after the ruling coalition split over whether to give citizens the right to vote in referenda.
Hmm, interesting one. Is the defined item 'PM Wim Kok's cabinet', 'the ruling coalition' or 'government(s)'- or all three??
The above examples all have 'one of' occurring consecutively, thus are more obviously defining text, but what about the following?
6) One woman member of Congress tried to commit suicide as passions mounted over Gandhi's exit, which came after three senior colleagues said her foreign origins and inexperience would hobble the party in elections later this year.
Pretty straightforward, 'Gandhi's exit' clarified by 'after three colleagues ... this year'. But what about 'One woman member of Congress' - defined by 'tried to commit suicide'?? Or am I getting carried away seeing definitions/clarifications everywhere?!
I've got six pages of this stuff, and that's only from one corpus and one search item. Anyone like to comment on my interpretations? Pinkie, none of your GE terminology, please! Of course, I can see loads of lexical-syntactic patterns and obvious signals such as 'which', etc, and in some cases, 'definitions' embedded in the text of other definitions, but I can't deal with hypotaxis (whatever that is) and all that other groovy stuff you GE-types are so attached to (OK, OK, I know I'm letting myself in for a slagging here).
In return for your (anticipated) analysis, I attach extracts from the MSc List corpus for 'empirical' and 'transitivity' - no occurrences of 'heuristic'. Maybe we ain't that sophisticated yet?!
Cheers,
Mary Lynn
Re: TDA & Defining text | Pinkie | May 16th, 2002
Regarding your TDA stuff:
I don't know whether I'm able to help, but it's fun trying. I'm going away today till Sunday, and have printed out a copy of your message to take with me! Would it be useful to focus on instances of "ONE OF" occurring after a comma, and similar types (e.g. WHICH, WHO, WHERE, WHEN after a comma; A, AN, THE after a comma)? Though I realize this is pre-judging the lexical realizations of the "defining/clarifying" function. It's a shame you can't do a concordance for commas: though certainly in word processors you can search for punctuation. Don't know if this is any use.
Best again
Pinkie
Re: Defining text | Mary Lynn Hughes | May 19th, 2002
Hi again Pinkie,
Many thanks for looking at this. I agree with you about 'relative' pronouns and commas (although of course commas aren't always involved), as well as 'a/an' and commas, etc (the use of brackets sometimes fulfills the same function). But I really can't start using Word to search for commas! (though it might be an option for particularly fine-grained analysis)
Another possibly hard-to-search-for feature I've noticed is repeated patterns (usually separated by commas) like:
• 'the ..., the ...' [can't find an example just now!]
• 'It VERB-s/-ed/-ing, it VERB-s/-ed/-ing...' [E.g. 'It is so real because it exists, it is there' - weird example, but you get the idea]
• two or more consecutive phrases/clauses starting with 'VERB-ing' [E.g. '...didn't feel up to one of Liz's 'rigorous' holidays (probably being cold and uncomfortable, getting little sleep and irregular meals, then coming back late and starting the week tired...']
• lexical repetition [e.g. '...all the newspaper reports, er reports of changes in air pressure affecting people's ears, er reports of leaks... there's a report here about how if there was a serious fire in one of the carriages....'; also interesting the repetition of 'er']
And other similar types of repeated clausal/phrasal structures, sometimes without the kind of lexical signals more conventionally associated with definitions/clarifications, e.g. 'this is', 'it's & it is/was' 'there is/are', 'is that' 'is what', 'the one', and all the relative pronouns, etc. But 'one of' is really rich in patterns and combinations of all kinds (I've now got about 15 pages to look at).
Anyway, lots of fun!
Cheers and hope you had a good weekend (with or without analysing my examples),
Mary Lynn
Re: Defining text | Colin Graham | May 21st, 2002
Hi Mary Lynn
OK here's my take:
I'm not sure what you mean by defining. To me a definition is defined by the use of "conventional defining formulae" such as are found in a dictionary, for example.
If by 'defining' you mean making a subset narrower or restricting it in some way, then it may be more helpful to identify the kinds of restrictions or identifications that are made.
I would suggest, for example: exemplifying, specifying, identifying, attributing, etc. as subsets of your larger role of 'defining'.
But:
'one of' is really rich in patterns and combinations of all kinds (I've now got about 15 pages to look at).
I'd suggest this and keep the other patterns as interesting questions for future research!
Your first 5 examples all have a pattern "one of the" whereas number 6 uses "one woman member of". Maybe 6 could be restated "one of the female members of Congress" to make it 'fit the pattern'. You could then look at the words following "one of the" and group them in subsets. For example, followed by a group "the greats, the world's..." or by a superlative "the most..., the shortest-lived..." and see what comes from that. From my experience with GE, I would suggest concentrating on one pattern or group of patterns in depth, and then looking for instances of it outside your corpora to check the validity of your findings. That's a kind of CL approach. You could also look at whether "one of the" pre- or post-modifies a noun, whether or not the noun group is a subject or object in the sentence... etc.
Have you got any concordancing software? It isn't difficult to use, you can download simple programs free and it makes searching for patterns so much easier if you have a concordance. I can provide further info or links if you are interested.
For clarifying, I think your task is easier. I would look for "mean by" "for example" "namely" "viz." and so on and look for the kind of things which are clarified, or specified. Is it new terminology? Does the clarification use a kind of defining terminology by analogy or example? Where does the 'clarification' come in the text/paragraph/sentence....?
Anyway, that's more than enough ramble from me for just now. Hope this helps rather than confuses!
Have phun!
Colin
:-)
GE/TDA: of-phrases, mostly | Pinkie | May 21st, 2002
Hi C, J, ML and anybody else who enjoys the linguisticky bits...
Mary Lynn: Interesting how your approach to this differs so dramatically from mine: must be all that mescaline you've been taking! You set out to identify how a function is achieved by examining texts; I'd already focused on particular lexicogrammatical realizations. I think your approach is probably better: though I admit to being a bit confused about exactly what you mean by "clarification." From the examples you give couldn't it be just about anything?! [How time flies: it doesn't seem like a year since you said you were off to Mexico!]
Signin' off,
Pinkie
Spain
Re: Defining text | Mary Lynn Hughes | May 21st, 2002
Pinkie,
Ha, ha. Mescaline me! (a little lexical test for you). Don't think I'll get into a lex vs. gram argument with you today, though I'll happily take up the sword again on some future occasion.
OK, by 'clarification' I mean what we do whenever we use a lexical item we think might not be clear to our reader/listener(s). And yes, from my tentative explorations, it could be almost anything- maybe that's what we're doing all the time, clarifying our message in one way or another. However, for TDA I have to narrow the focus, so very roughly, text that can be seen as clarifying the (intended) meaning of some specifically-mentioned lexical item. In my investigation, I'm not specifically targeting differences between 'defining' text (as distinguished from formal definitions) and clarifications, although intuitively, there are differences and these may become clearer along the way (probably with some area of overlap).
One potentially interesting angle is whether the 'defining' text more typically precedes or follows the lexical item. So far, it seems it can be either, but this may depend on some factor I haven't identified (maybe there are typical sequences for different types of lexis, different discourses...). Because I'm interested in how 'definitions'/clarifications are pragmatically effected, I've left the potential lexical realisations wide open, at least initially. It's easy enough to follow up the obvious ones, but more interesting I think, and potentially more rewarding(?) to look at some of the less common/well-known ones that appear in the data.
But, I'd love to have your interpretation of my 'one + of' examples (16 May post) and the other patterns I mentioned yesterday. (The 'one of' ones should be right up your alley, as per your 'part of the city' etc examples).
Best wishes,
Mary Lynn
Re: Defining text | Pinkie | May 21st, 2002
Hi Mary Lynn:
Here goes then... In my opinion, none of your sentences 1-6 (pasted below) can be reasonably said to contain elements with clarifying function as you define it. Three of the sentences contain clauses which provide additional information about something already mentioned: "in education..." in sentence 2, "regarded..." in sentence 3, and "who died..." in sentence 4. But "providing additional information" is such a broad category that it's pretty much meaningless and additional to what? How do you decide which information is "core" and which "additional"?
As I understand it, clarifying function as you (very reasonably) define it is realized in
"A Swisstool. You know, one of those next-generation pocket knives with pliers, saws, knives, etc."
(The speaker presumably suspects hearer may not understand "Swisstool", so clarifies).
A similar example, though this time for humorous effect, is:
"Especially if you drive an actual car - you know, one of those old-fashioned, pre-SUV vehicles that DON'T have eight feet of clearance underneath."
Related but not the same:
"As she hit the floor her blouse rode up, you know, one of those man-made fibre chiffon things."
(this is not really a clarification of "blouse" in the same way that the first example contains a clarification of "Swisstool").
Anyway, like I say, I don't think any of your sentences contain elements that realize functions of this type, despite the fact that they all contain "one of". Again, I think a feasible way of locating this type of use of "one of" is to focus on ", one of" (i.e. occurrences preceded by a comma).
Anyway, maybe I'm not understanding you right: please don't take me very seriously.
"Mescaline me ---"? I give up!
Best,
Pinkie
Re: Defining text (reply to Colin) | Mary Lynn Hughes | May 22nd, 2002
Colin,
To clarify what I'm doing: It's precisely the non-conventional 'defining'/clarifying formulae I'm interested in and how these are signaled within text, rather than dictionary-type definitions. I'm looking to identify lexical items and lexical-syntactic patterns commonly used for this purpose (not just the more obvious ones like 'means', 'namely', 'called' etc). The 'one of' pattern happens to be one of(!) the first I've looked into in detail. (And thanks for your observations of the patterns that follow it- they are similar to what I noted). It's a corpus-based investigation, and yes, I'm using the ATA software. Also, I'm (mostly) interested in how this occurs in 'general' discourse, rather than in specialised/restricted domains- what strategies are used for 'defining' or clarifying lexical items in a variety of text types (including spoken). I'm using three corpora: newspaper text, CAE reading & listening texts and email messages. (I explained this in my post of 16 th May, maybe not very clearly!).
I started by reading five or six individual texts and noting down the 'defining' language used in them. This revealed a number of common features, including the ones you mentioned- a variety of ways of exemplifying, identifying, listing attributes, etc. Probably the most common pattern is 'is a/an ...', but there are many many others, such as 'is the ...', 'the -est ...', 'such a ...', '(not) as ... (as)', to mention just a few. In addition, there are recurring lexical items that might be worth investigating, some fairly predictable ('uses/used/using', 'like', 'represent(s)', 'find/found', 'seem(s)', 'provide(s)', 'way(s)', = 'type', 'kind', etc); some less so ('apparent(ly)', 'says', 'really', 'thing', and even 'just').
Another aspect is cohesive ties (esp all manner of pronouns, deixis, substitution, lexical repetition, etc) which add information to the 'definition' but may not be in the immediate vicinity of either the item defined or the main 'defining' text. (Quite a few examples of all this in this message! My whole first paragraph could be considered a definition/clarification, and this one is full of examples and exemplifying lexis.) All very fascinating, but I can't investigate everything, so am selecting a few main patterns, a few less obvious ones and some lexical items to follow up, using the corpus software to look at concordances and sentences- and where necessary, going back to the original text to examine preceding/following text. Even this is time-consuming, so don't know how far I'll be able to take it.
Interestingly, there's a fair bit of research into textual definitions of specialised terminology, but not any(?) as far as I know on 'defining' text in more general discourse (hence, a problem in terms of references- IF ANYONE KNOWS OF ANYTHING, I'D LOVE TO HEAR ABOUT IT). BTW, there's a good example of a conventional defining formula in the TDA module (Unit 5, pg 7). I decided on this topic for TDA partly because it might be useful to raise learners' awareness of how definitions/clarifications are effected in text (e.g. for reading comprehension), but especially because it feeds into a larger project on how meaning- especially the variety of meanings associated with a lexical item- might be identifiable through looking at corpus extracts containing the target item. So that's it, if you've managed to make it this far!
Time to stop yakking and get back to work!
Cheers,
Mary Lynn
Re: Defining text (Reply to Pinkie) | Mary Lynn Hughes | May 22nd, 2002
Pinkie,
Thanks mucho for your response. I love your 'clarification' examples- are they authentic?! Results of a corpus search for 'you know'? I would guess that's one of the common lex items for this purpose, along with 'I mean'- both perhaps more likely to occur in spoken or informal written text, cf 'X means...', one of (!) the key lex signals in 'definitions' generally.
However, I hadn't intended the six extracts in my 16 th May post as CLARIFICATIONS; they were just my first shot at analysing some of the 'one + of' data (plus any other 'defining' text I spotted in the extracts). I see them as 'defining' (as per my interpretations) and wondered if others would agree/disagree with what I identified as the 'defined' and the 'defining' text, since I wasn't always sure about it.
Fasten your seatbelt or eject now! What follows is not for the faint-hearted.
Re: Additional information.
You wrote:
Three of the sentences contain clauses which provide additional information about something already mentioned: "in education..." in sentence 2, "regarded..." in sentence 3, and "who = died..." in sentence 4. But "providing additional information" is such a broad category that it's pretty much meaningless; and additional to what? How do you decide which information is "core" and which "additional"?
[For anyone who's insane enough to try to follow this discussion, Sentences 2, 3 and 4 are reproduced at the end of this message, along with my interpretations]
Good point, although defining what's 'core' and what 'additional' is not key to my investigation, which simply hopes to identify (some of the) lexical signals for what may be seen as 'defining' text. Which is not to say I oughtn't at least look at some of these 'messy' aspects. In my 'reply to Colin' post today, I mentioned that the additional info 'may not be in the immediate vicinity of either the item defined or the main 'defining' text'. Presumably, it will come at some point after the 'main defining text'? Agreed that the latter may not always be easy to identify and may well necessitate looking at more of the co-text to see whether or not these 'one of' clauses are adding info to something already 'defined'.
I've just done that for Sentence Two:
2) 'With ONE OF THE MOST UNDERDEVELOPED MUSICAL INFRASTRUCTURES IN
EUROPE [...] this State has yet to define or declare a national music policy.'
What I found (in the second paragraph of the article, much earlier than Sentence
Two) was
A) 'The employment of a single small group of musicians may seem a small move, but in terms of THE VERY LIMITED RESOURCING OF MUSICAL ACTIVITY IN THE STATE it is of great significance.' [My caps in both quotes] Where, I think 'the very limited resourcing of musical activity in the State' could be the 'defining text' that the 'one of' clause in Sentence Two is 'adding info' to. Especially if you take into account the text immediately preceding Sentence Two, AND the final sentence of the article, which immediately follows it- all, including Sentence Two in [] following:
B) 'The PIANO Report, which was so important in vaulting IAPA onto the political agenda, also pointed out that there is no official school of music north of a line from Dublin to Galway. Norway, with a population of 4.2 million, has more than 350 music schools and growing. Iceland, with a population less than County Cork, has over 70. Welcome as both the IAPA and Sligo quartet projects are, they are actually isolated developments. [With one of the most underdeveloped musical infrastructures in Europe- in education, Northern Ireland is luxuriantly resourced by comparison- this State has yet to define or declare a national music policy] For the major agencies involved- RTE, the Arts Council, the Department of Education and the Department of Arts and Heritage- it is a vacuum demanding to be filled.'
The Norway and Iceland examples provide supporting evidence for 'the very limited resourcing...' in A (though unhelpfully for my lexical purposes, they're not explicitly signaled as examples), reinforced by 'isolated developments'; all of this is summarised by Sentence Two's 'one of the most underdeveloped...'; and clinched with 'a vacuum demanding to be filled' in the concluding sentence.
So, going back to the original question, can I say that 'ONE OF THE MOST UNDERDEVELOPED MUSICAL INFRASTRUCTURES IN EUROPE' in Sentence Two provides 'additional info' for the (earlier) 'defining' text in A: '(in terms of) THE VERY LIMITED RESOURCING OF MUSICAL ACTIVITY IN THE STATE')? Would you agree that this is in fact 'defining text'- for 'is of great significance'??
Gawd! This is all horribly convoluted and difficult to follow. And I've spent far too much time on it. Think I'd better withdraw to quiet contemplation of my data and leave the hurly-burly of the list alone for a while. Still, I'll be glad for any responses, though I can't guarantee to answer them (doesn't make responding seem worthwhile, does it?). Pinkie, many thanks. And Colin too- you've both made me think harder about my interpretations of 'one of' and what I'm doing in general.
Cheers,
Mary Lynn
PS: 'arse'- at least in my idiolect.
Re: Defining text (Reply to ML) | Colin Graham | May 26th, 2002
Hi ML,
It's a corpus-based investigation, and yes, I'm using the ATA software.
Can you tag your corpus for cohesive ties? Maybe using the cohesive ties idea from Halliday and Hasan (p329 on) would be helpful. Or Hoey's matrix ideas in Patterns of Lexis in Text.
Interestingly, there's a fair bit of research into textual definitions of specialised terminology, but not any(?) as far as I know on 'defining' text in more general discourse (hence, a problem in terms of references- IF ANYONE KNOWS OF ANYTHING, I'D LOVE TO HEAR ABOUT IT).
I have a number of corpus-based ELT books. Most of them are published by Benjamins. You might find these ones of most use:
Partington, A (1998) Patterns and Meanings. Amsterdam: John Benjamins.
Ghadessy, M., Henry, A., Roseberry, R. L. (2001) Small Corpus Studies and ELT. Amsterdam: John Benjamins.
Maybe:
Garside, R., Leech, G., McEnery, A. (1997) Corpus Annotation. Harlow: Addison Wesley Longman.
Or:
Stubbs, M. (2001) Words And Phrases. Oxford: Blackwell
The last two may be of more general interest, although Stubbs suggests using frequency lists of texts to look at keywords and their distribution that is quite interesting.
Partington has some nicely grouped examples of using corpora to investigate metaphor, syntax, cohesion, etc.
Ghadessy et al has a paper by Barnbrook and Sinclair that looks at local and functional grammars. More specifically how the defining language used in the COBUILD dictionary has its own special grammar. There's also an interesting paper by Ragan which uses a systemic functional corpus in the classroom- maybe your long-term aim for the defining language?
If you can tag your corpora, then you may also find it helpful to look at tagging synonyms, hyponyms, superordinates, complementaries and so on. Which is the most popular method of clarifying....?
Mr Blair, the Prime Minister.... (Synonymous at the time of writing)
Member of Congress... Parliament (hyponyms of systems of government)
etc. If you limit yourself to one or two main patterns, as you suggest, then there may be an interesting usage of the "nyms" that appears. Is 'one of the' hyponymous or synonymous, for example.
Anyway, have fun with the reading!!
BFN
Colin
:-)
Re: Defining texts | Mary Lynn Hughes | May 26th, 2002
Colin,
Many thanks for your reply. I moved to a new house last weekend and have just got operational again.
I don't think there's a facility for tagging with the ATA software, so I can't go down that road. Pity, it would make identification easier/more systematic, though presumably it's a fair bit of work to set up the tagging?
Thanks a lot for the refs. I'll see if any of them are in my uni library. I've read a little about the defining language used by Cobuild, but again, it's at the more predictable end of the continuum and not what I'm (mainly) looking for. Don't think I mentioned it, but the book I've found most useful for my purposes is: Pearson, Jennifer, 1998. Terms in Context. Benjamins. Although she's looking at mechanisms for identifying definitions of technical terms in highly specialised discourses, mainly for translation purposes, she outlines a wide range of defining stategies and uses a completely corpus-based approach. Might be of interest to some of you EAPers out there?
All your suggestions are most interesting. The main problem is not having time to go into all this in nearly as much depth as I'd like to. In another, post-MSc life, perhaps?
I must have stumped Pinkie completely with my last OTT message, as he hasn't replied. Just call it assignment pressure hysteria (on my part, I mean).
Best to all,
Mary Lynn
Re: Defining text | Pinkie | May 31st, 2002
Hi Mary Lynn..
I must have stumped Pinkie completely with my last OTT message, as he hasn't replied.
Now there's an interesting discourse feature: I think we might call that "challenging" text!
I got my clarification examples by searching for "you know, one of" in the exact phrase box of Google Advanced Search; first change preferences to show 100 hits at a time, and to turn off the family filter. By the way, have you ever used the ICE corpus? It's titchy (1 million words), but very nicely tagged, and allows you to search for e.g. NOUN+NOUN. It's also very comprehensively categorized by text type.
Getting to the main point, you wrote...
'(in terms of) THE VERY LIMITED RESOURCING OF MUSICAL ACTIVITY IN THE STATE')? Would you agree that this is in fact 'defining text' - for 'is of great significance'??
No, I wouldn't, sorry...
1) Occasional exceptions notwithstanding, I think that you can only "define" a noun phrase (or the thing represented by that noun phrase, if you prefer). You can't meaningfully "define" a clause fragment like "is of great significance".
2) I think the text features you're pointing to are best approached as examples of the very general issue of information structuring rather than of the very specific act of defining. Basically you're talking about first mentions of information and subsequent mentions. [Though interestingly your Sentence Two, which apparently contains information closely related to the first mention of "the very limited resourcing...", contains no textual signals to indicate that this is the case: in other words, the information is presented as New.] In my opinion, then, EITHER you're talking about defining, in which case your examples aren't appropriate, OR you're talking in rather general terms about Given/New structuring. As ever, though, I may simply be missing what you're getting at.
Thanks for the mention of Pearson "Terms in Context": I'll see if that's in my library.
Hope the new house is up to standard!
Best,
Pinkie
with 
and 
