Corpus Investigation of the phrase 'Public Interest'

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

CENTRE FOR APPLIED LINGUISTICS

MA ASSIGNMENT COVER SHEET

STUDENT ID NUMBER: 1163612

PROGRAMME: MA IN ELTMM/ICT

MODULE NAME: LEXIS AND GRAMMAR – ET904-B

MODULE TUTOR: SUE WHARTON

WORD COUNT: 2,195

ASSIGNMENT QUESTION:

DATE DUE: 12 NOON on 11 JANUARY 2012

DATE SUBMITTED: 9 JANUARY 2012

In completing the details on this cover sheet and submitting the assignment, you are doing so on the basis that this assignment is all your own work and that you have not borrowed or failed to acknowledge anyone else’s work

Please X this box if you agree to this statement X

0

Assignment B

Using lexical analysis software, investigate the behaviour of a word or word sequence in either a large general corpus, or a smaller specialised corpus. Discuss the collocations, semantic associations, and grammar patterns of your chosen word or word sequence.

Your assignment should include (word counts are given as guidelines only):

- A brief discussion of your reasons for focusing on your chosen word(s), of the corpus on which your analysis is based, and the analysis procedures you undertook (100-200 words)

- Your analysis of the words’ behaviour in the corpus

- A discussion of the significance of your analysis, explaining what you have discovered about the behaviour of the word, either generally or within a specialised domain.

Concordance lines, tables of collocates etc. may be included as appendices.

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

An analysis and comparison of the behaviour of the phrase,

‘Public Interest’

using a large general corpus and a subcorpus of UK newspapers.

INTRODUCTION:

The phrase, ‘public interest’, has a long history, with a current topical significance i. I am interested in

knowing whether lexical priming can account for the way in which this phrase is used. Using

Hallidayan references, Hoey’s (2005) theory, and a large general corpus, the Collins Wordbank ii, I will

investigate the behaviour of the phrase.

‘Public interest’ can be defined simply as ‘the common welfare of the people’. It can also be defined

more broadly as:

a common concern among citizens in the management and affairs of local, state, and

national government. It does not mean mere curiosity but is a broad term that refers to the body

politic and the public weal. A public utility is regulated in the public interest because private

individuals rely on such a company for vital services.iii

This definition is concerned with the fact that the people have a right to know something because it

may affect them directly. However, there is a more ambiguous definition such as:

the fact that people in general are interested in something.iv

A publicised event can be cancelled, for example, because of a lack of ‘public interest’.

Analysis – collocation:

The individual words ‘public’ and ‘interest’ are already widely used together. To make any claims

about what other words ‘public interest’ collocates with, we need to use corpus evidence. We can

define collocation in statistical terms as the sequences of words which occur more often than would

be expected by chance. (Sinclair, 1991). Similarly:

If words commonly occur in the same text and we are frequently exposed to their co-

occurrence, we come to expect them together (Bloor and Bloor, 1995:101).

A search for the phrase ‘public interest’, across all text types, generates 2517 hits [see Appendix A].

A further search for collocates of ‘public interest’, using standard settings v, sorted by T-scorevi, brings

1

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

up the definitive article, ‘the’ (Frequency: 2305; T-score: 45.760), and preposition, ‘in’ (F: 1220; T:

33.848), as the most common. Adjusting the concordance horizon range to -5 to 0, to show just

those collocates to the left of the phrase, results in (F: 1710; T: 38.739) and (F: 758; T: 26.161)

respectively [Appendix B]. If the T-score is ‘the confidence with which we can assert that there is an

association’ (Collins, 2009, release notes) then we could readily claim that ‘the public interest’ and

the prepositional phrase, ‘in the public interest’, are the most common collocates. This adds a

significant meaning towards the concept of ‘public interest’ as it implies that an action, such as a

disclosure, is carried out for the benefit of the public, adding a general appeal, relevance or

importance.

The high frequency of a full stop (F: 1078; T:29.500) and the capitalised, definitive article, ‘The’ (F:

227 14.120), suggests that the phrase appears commonly at the end of a sentence. Further

investigation shows this is generally true, with substantial appearance at the beginning of a sentence

also [Appendix C]. The words ‘Research’ (F:52) and ‘Group’ (F:50) both capitalised, appear

exclusively in US literature, as part of the wider phrase, ‘Public Interest Research Group’ [Appendix

D].

Using ‘word sketch’vii, the word ‘public’ acts as a modifying adjective for the word ‘interest’ 1952

times [Appendix E]. This is ranked fifth after ‘service’, ‘opinion’, ‘school’ and ‘sector’, which are

effectively more common collocates. By comparison, ‘interest’, as a noun, is modified by ‘public’

more than any other word.

Narrowing the focus using ‘Newspapers (form) > news (domain) > UK (country)’,viii this filtered

search, or ‘subcorpus’, generates 647 hits. Again, ‘in’ (F:334; T:17.745) and ‘the’ (F:605; T:23.468)

have the highest frequency [Appendix F]. 647 represents more than half of the hits (1016), found if

filtering just by country and this reflects the high percentage of newspapers (51.80%) in the corpus.

An analysis of the subcorpus reveals the following frequencies of ‘parts of speech’, within the

concordance horizon of -5 preceding the phrase, ‘public interest’:

Determiner (the, a) 547 hits [Appendix G]

Preposition (e.g. against, for, in, under) 479 [Appendix H]

Noun (e.g. Government, consideration) 387 [Appendix I]

Verb inc. Modals (e.g. is, was, will, could) 295 [Appendix J]

Pronoun (e.g. it, who, he) 135 [Appendix K]

2

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

Within the subcorpus, the phrase is preceded by ‘the’ 350 times, and by ‘in the’ on 190 occasions

[Appendix L]. Just as ‘in the’ collocates highly with ‘public interest’, the verb, ‘to be’, in some form or

other, is the most commonly collocated verb with either version. Thus, the lexical chunk, ‘is (not) in

the public interest’ is widely used. ‘Being’ can also be described as a ‘process’, which I will return to

when looking at colligation.

Priming as an explanation of collocation

Hoey (2005) states that the widespread use of collocations can be explained by ‘priming’, which is a

psychological concept. We can only account for collocation, he argues, if every word is mentally

primed, or preferred for collocational use. As a word, or phrase:

is acquired through encounters with it in speech and writing, it becomes cumulatively

loaded with the contexts and co-texts in which it is encountered, and our knowledge of it includes

the fact that it co-occurs with other words in certain kinds of context (Hoey, 2005:8).

If members of the public were asked what word succeeds ‘public’, they would not necessarily say

‘interest’ (F:1952; T: 7.87), but they might. ‘School’ (F: 3517; T: 8.28) is arguably more likely to be

the response. It is also arguably more likely that asked what word precedes ‘interest’, a higher

proportion would say ‘public’ than most other words, such as ‘commercial’ (F:417; T:7.6).

Stubbs (1996, in Hoey, 2005:8) states that:

‘Speakers are free, but only within constraints … the reproduction of the system is the

unintended product of [a speaker’s] routine behaviour’.

However, can we say that the meaning of ‘public interest’ has been primed over long exposure to

the phrase? Is there a constraint to how it is used? Furthermore, if priming leads to a speaker

unintentionally reproducing an aspect of language which subsequently primes the hearer (Hoey,

2005:9), then can the meaning associated with that phrase change in the way that definitions of

some words change over time? I will return to this later.

Semantic Association

Whatever definitions are given to ‘public interest’, the semantic associations are quite separate from

the two words that it combines. Hoey (ibid:12) states that if priming can account for collocation,

then it opens up that priming can explain other features of language. One feature is the semantic

prosody (positive or negative connotations) of a phrase. It is arguable that the only positive or

negative connotation of our phrase is whether something is or is not deemed ‘to be of it’. A piece of

3

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

information, or the act of uncovering one, can be expressed in terms of whether it is or isn’t in the

‘public interest’, but this is more clear cut, I would argue, than suggesting that connotation exists.

Semantic preference (Sinclair, 1991) or semantic association (Hoey, 2005) are terms used:

when a word or word sequence is associated in the mind of a language user with a semantic

set or class, some members of which are also collocates for that user (Hoey, 2005:24)

A list of semantic associations is difficult to draw up. It is easier to focus on ‘parts of speech’. The

original corpus contained a high frequency of short, prepositional words [Appendix B]:

Preposition Frequency

Range: -5 to 5

T-score/Ranking

(excluding

punctuation)

Frequency

Range: -5 to 0

T-score/Ranking

(excluding

punctuation)

Of 640 23.172 – 3rd 459 18.914 – 3rd

To 621 22.674 – 4th 348 15.654 – 5th

For 317 16.760 – 8th 221 13.615 – 6th

By 138 10.970 – 17th 50 5.780 – 26th

On 107 8.857 – 25th 61 5.840 – 24th

Against 80 8.762 – 27th 68 8.048 – 17th

Adjectives which modify the phrase appear less commonly than one might predict. The first one,

‘intense’ is found in 65th place – (F:29; T: 5.372), with ‘contrary’ second in

67th (F:28; T: 5.285). The third, ‘legitimate’ (F:24; T: 4.891) forms part of a

small, ‘legal’ noun set (law, legitimate, litigation, judge, lawyers) [Appendix

M]. Other semantically associated words, such as ‘Government’ and

‘political’ are found, but not in significant numbers, according to this data.

Premodifiers, for example, adjectives that can be grouped by size

(‘great(er)’ F:34; ‘growing’ F:11; ‘huge’ F:11; ‘enormous’ F:6), or grouped

by gravity (‘broad’ F:16; ‘significant’ F:5; ‘outstanding F:1) are even less

apparent. ‘Public interest’ is, therefore, a phrase which somewhat

surprisingly, according to this data, only occasionally gets preceded by a

word which pre-modifies it.

4

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

Other terms appear, but not as frequently as might be predicted. For example, the words, ‘attract’ in

different ‘lemmatized’ versions (F:17), (Appendix N), ‘clearly’ (F:15), (Appendix O) and the phrase,

‘lack of’ (F:12) [Appendix P].

Pragmatic Association

Does the phrase attract what Hoey (2005) terms pragmatic association?

[This] occurs when a word or word sequence is associated with a set of features that all

serve the same or similar pragmatic functions … The boundaries between pragmatic association and

semantic association are not going to be clear cut, because priming occurs without reference to

theoretical distinctions of this sort (Hoey, 2005:26).

For there to be pragmatic association here, the phrase, ‘public interest’, would need to be shown to

be associated with words which have a similar pragmatic function.

A search using the subcorpa and the collocation tool generates a word, ‘ Immunity (F:8 – 1st) /

immunity’ (F:18 – 4th), at the top, when ranked by MI score (see also note v). The word ‘immunity’

appears because of a wider legal context of a witness being protected and, therefore, can be

associated with the wider semantic association mentioned earlier, along with ‘prosecuting’,

‘evidence’, ‘court’, ‘Appeal’ and the lemma, ‘disclosure’, which appears in three forms, with the

highest-ranked one due to ‘the Public Interest Disclosure Act 1998’.

However, in searching for pragmatic functions, I seem to have discovered another list of semantic

associations, albeit in a different way. Would I have to find words with a similar pragmatic function

before rejecting the notion that priming is taking place? ix

Colligation

Just as a lexical item may be primed, or preferred, to co-occur with another lexical item, so also ‘it

may be primed to occur in or with a particular grammatical function’ (Hoey, 2005:43). One way of

analysing the grammatical behaviour of the phrase is using the KWIC sentence view [Appendix Q],

rather than viewing the node phrase within the context of words either side.

Immediately, it appears that ‘public interest’ prefers a complement position, either as direct object -

‘The only moralists (Subject) who (relative pronoun) have (Finite) attracted (predicator) public

interest (Cdo)…’ – or as part of an intensive complement - ‘Largest chains are abusing their market

power against the public interest (Cint)’.

5

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

Rather than being the subject, as you would expect if an article was about ‘(the) public interest’, the

phrase appears, more likely, to be the direct or indirect complement of some other action. It is not

frequently used as the head of a noun group. This, in turn, fits with the original discovery that ‘ in

the’ was the most common collocation to the left of the phrase. We could suggest, therefore, that

the phrase is positioned towards the end of a clause or sentence, as an external reason for an action

taking place. A further example:

The prepositional phrase, ‘in the public interest’, could be seen, however, as a kind of ‘exophoric’,

circumstantial adjunct, rather than a complement. It does not answer the question, ‘who or what?’

in the traditional complement sense (Bloor and Bloor, 1995:47). Nor is it something which has an

action done to it. But it performs a ‘slightly peripheral’ function in the clause (ibid:51). It is

grammatically optional to include it, as the sentence above could have ended with ‘… a jury who

accepted his defence.’ But it does provide an external reason –hence the ‘exophoric’ label - for an

action or process, ‘being acquitted’. Similarly, it could be a dependent factor on an action taking

place in the future:

The ‘public interest’ is a general concept of something which exists, is generally understood, but

which has contrasting definitions. It could only be a ‘participant’, that is, an element in the

ideational function of a clause, during a discussion about its definition, such as the example below:

Transitivity

Earlier I noted the frequency of the verb ‘be’ used with the phrase. ‘Be’ is part of a verbal group,

made up of lexical verbs (open set) and auxiliary verbs (closed set), which is frequently represented

in clauses using the phrase. The Hallidayan concept of transitivity includes the function of process

types, as explained:

The process centres on that part of the clause that is realised by the verbal group, but it can

also be regarded as what goings-on are represented in the whole clause (Bloor and Bloor, 1995:120).

6

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

The phrase does appear to attract what can be described as ‘relational process’ types, although this

also contains difficulties:

The semantics of relational process is very complicated and different sets of Participant roles

can be associated with different, more delicate categories of Relational Process. Out of context it is

often difficult, frequently impossible, to subclassify Relational Processes precisely (ibid:110).

Relational processes are, put simply, ‘being’ or ‘having’ verbs. Out of 647 concordance lines, 370

include a form of ‘to be’, while 66 include a form of ‘to have’ [Appendix R]. From this, we can

continue to build up collocated expressions, such as ‘to be in the public interest’. I earlier suggested

that the prepositional phrase, ‘in the public interest’, is a kind of circumstantial adjunct. This term is

usually reserved for showing when (time) and where (place) an action ‘goes on’. In an abstract sense

it tells us where the relational processes are taking place. When an action takes place it is frequently

said to ‘have’ or ‘be’ in the public interest. Furthermore, in Hallidayan SFPCA functional terms,

relational processes frequently take an intensive complement, which fits the pattern of many

sentences featuring the phrase.

Discussion:

I began with some definitions of ‘public interest’. One frames the concept as having direct relevance

or importance, whilst the other, more ambiguous term, refers to a general interest. There is no UK

parliamentary definition, although meaning is often constructed vis-à-vis the ‘substantial harm’ test

or the ‘freedom of information’x. I am interested in the how certain sections of the print media have

redefined what is ‘in the public interest’ as information quite removed from the original definition of

people having a right to know something because it affects them directly.

How the phrase behaves in this particular corpus, however, does not appear to change depending

on the definition used. The phrase seems to act significantly as an exophoric, circumstantial adjunct

as part of a relational process. It appears to position itself more towards the end of a sentence or

clause than at the beginning and forms a wider context for more specific actions to take place.

I earlier asked whether priming is involved in the way the phrase is used. It is possible, I would

suggest, that long exposure to the idea that something of general interest is ‘in the public interest’

has skewed the original meaning somewhat. The meaning associated with the phrase may have

changed over time. ‘Priming’ operates on a psychological, unintentional level and is concerned with

unintentional reproduction. This does not appear, however, to be the case with this phrase. It might

7

ET904-B: LEXIS AND GRAMMAR STUDENT ID: xxxxxxxx

have certain behavioural characteristics, but it is questionable whether its behaviour is affected by

priming. That is not to say that priming does not take place for other lexis, which clearly it does.

8

i The phrase, ‘in the public interest’, is frequently associated with the justifications of media investigations and this is a topical issue in the UK, with the general ethics of the printed press, currently being investigated under the Leveson Inquiry. This inquiry includes, amongst other ‘modules’, an inquiry into the relationships between the press and police and the extent to which that has operated in the public interest. It is a ‘slippery concept’ and has been widely discussed at the ongoing inquiry (see Appendix S for fuller quotes).

There is currently no parliamentary definition of ‘public interest’. Its definition is highly important as it used in framing concerns over the justification for certain practices by the press, including concerns on specific matters by the chairman:

I do entirely understand the significance of the issue and I recognise that it is likely to be in the public interest that this be resolved in an orderly manner.

Lord Justice Leveson, 14 December 2011

The issue that Lord Justice Leveson is referring to in this particular quote is whether the Guardian newspaper was accurate to describe ‘as fact’ that the News of the World had deliberately deleted voice messages from the mobile phone of the murdered schoolgirl, Amanda ‘Milly’ Dowler.

Ideally, an analysis of a corpus which includes all UK newspapers, not a selection may produce different interpretations of the phrase ‘public interest’ in order to justify its actions, but may not produce differences in phrasal behaviour.

The way one politically oriented newspaper might be ‘primed’ to use the phrase may differ from the way a differently oriented newspaper might be ‘primed’. If, by primed, we are concerned by the psychological approach to the use of lexis, then surely this can include motivations, restrictions and coercions on the part of a newspaper owner to the writer, say, an editor or a journalist. Indeed, my own selections and interpretations might be influenced by the ‘priming’ that I have been exposed to.

ii The Collins Wordbank is a large general corpus, with 22,620 documents containing 553,171,489 tokens. 51.80% of the text comes from newspapers, 27.54% from books, 11.12% from spoken sources and 7.91% from magazines. Similarly, 52.31% of the texts relate to news. Source: Release Notes 2009.

iii The Free Dictionary http://legal-dictionary.thefreedictionary.com/Public+Interest

iv Macmillan Dictionary http://www.macmillandictionary.com/dictionary/british/public-interest

v Standard settings are to search for a key word in the range: -5 to 5, minimum frequency in corpus: 5, minimum frequency in given range: 3, by T-score and MI. Source: Release Notes 2009.

vi The Mutual Information score expresses the extent to which observed frequency of co-occurrence differs from what we would expect (statistically speaking). In statistically pure terms this is a measure of the strength of association between words x and y. In a given finite corpus MI is calculated on the basis of the number of times you observed the pair together versus the number of times you saw the pair separately.

MI does not work well with very low frequencies - the t-score provides a way of getting away from this problem as it also take frequencies into account. The t-score is a measure not of the strength of association but the confidence with which we can assert that there is an association. MI is more likely to give high scores to totally fixed phrases whereas t-score will yield significant collocates that occur relatively frequently. In most cases, t-score is the most reliable measurement.

vii Word Sketch shows the words which typically combine with a chosen search term, and what their grammatical relation is to the search term.

viii For the purpose of this subcorpus (of 122,719,932 tokens), the results include the Independent, The Times, The Sunday Times, The Sun and, interestingly, The News Of The World. It also includes four regional newspapers - Glasgow Herald, Belfast Telegraph, Irish Times and Liverpool News. Source: Release Notes 2009.

ix An attempt to carry out more specific searches for grammar, such as using the CQL (Coprpus Query Language) box were thwarted by an software system error. Screenshot below.

x

Amongst other relevant pages on the parliamentary website, this committee, set up in 2011, has been set up to look at the issue of defamation. Clause 2 asks the general public whether ‘public interest’ should be defined. http://www.parliament.uk/business/committees/committees-a-z/joint-select/draft-defamation-bill1/news/call-for-evidence/

REFERENCES:

BBC Newsnight: http://www.bbc.co.uk/iplayer/episode/b018b9kf/Newsnight_14_12_2011/ . Accessed on 14 December 2011.

Bloor, T and Bloor, M. (1995) The Functional Analysis of English: A Hallidayan Approach. London: Arnold

Davis, N. (2011) at the Leveson Inquiry: http://www.guardian.co.uk/media/blog/2011/nov/29/leveson-inquiry-nick-davies-paul-mcmullan-live#block-50. Accessed 17 December 2011.

Leveson Inquiry: Culture, Practice and Ethics of the Press: http://www.levesoninquiry.org.uk/

Collins WordbanksOnline (2009): ‘Corpus Concordance Sampler’ Available at: http://wordbanks.harpercollins.co.uk

Hoey, M (2005) Lexical Priming: A New Theory of words and language. London: Routledge.

Sinclair, J.M. (1991) Corpus, Concordance, Collocation in Hoey, M (2005) Lexical Priming: A New Theory of words and language. London: Routledge.

Stubbs, M. (1996) Text and Corpus Analysis in Hoey, M (2005) Lexical Priming: A New Theory of words and language. London: Routledge.

Appendix A:

First page only (20 concordance lines) of an unfiltered search for ‘public interest’, across the whole corpus, with default settings.

Appendix B:

In the range: -5 to 5 In the range: -5 to 0 (collocating on the left)

Appendix C:

Page 3 of 12 highlighting the collocate ‘The’ (capital letter) to show how the phrase ‘public interest’ commonly appears at the beginning and end of a sentence.

Appendix D:

Exclusively US literature found for the words ‘Research’ and ‘Group’.

Appendix E:

Grammatical relationship of ‘public’ and ‘interest’

Appendix F:

‘public interest’ filtered subcorpus: - Page 1 of 33

most popular collocates in the subcorpus.

Appendix G:

Appendix H:

Appendix I:

Appendix J:

Appendix K:

Appendix L:

Appendix M:

Highlighted results from whole, unfiltered corpus:

Appendix N:

Appendix O:

Appendix P:

Appendix Q:

Appendix R:

Appendix S:

Extracts which quote the phrase, ‘public interest’ during one day of hearing at the Leveson Inquiry, 29 November 2012. Available at: http://www.levesoninquiry.org.uk/hearing/2011-11-29am/

and at: http://www.guardian.co.uk/media/blog/2011/nov/29/leveson-inquiry-nick-davies-paul-mcmullan-live#block-50

One view:

12.01pm: Nick Davies, of the Guardian, says in paragraph six of his statement that the concept of public interest can be particularly "slippery". Expanding on this, he says in operational terms it's difficult because it's hard to tell where the lines are supposed to lie. The answer is it would be all right if it's in the public interest but we are stymied … very often it isn't clear and personally I would like it if somebody [set up] a public interest advisory body that I, or a member of the public, could go to and could get high quality advice.

In the advent of a dispute I would be able to provide the advice and say 'this is what I was told'. I would have something that was weighty in the event of a dispute.

12.03pm: In relation to the public interest, Davies says: I profoundly disagree that the News of the World had a public interest in publishing the story about Max Mosley's sex life.

12.39pm: Davies' written statement expands on the reasons why he felt it would be a breach of privacy to publish a story about a former minister's phone being hacked. The raw material for that story included details of messages which had been exchanged between him and a woman friend. I argued that we should not publish those messages - they were intrusive, and it was perfectly possible to expose the important point, that this minister had been a victim, without breaching his privacy.

The same kind of balance was raised by the story of the hacking of Milly Dowler's voicemail which I brought in in July 2011. I was sure that it was a matter of public interest that should be revealed, but I had some concern that publication would breach the Dowler family's privacy by exposing them to yet more publicity.

2.38pm: Davies is now discussing a story about David Blunkett's alleged relationship with a married woman. He says he would not have published this because it was "prurient" and an "unjust invasion into his private life"".However, as the journalists "dug in" to the story they did uncover something of public interest, which was that Blunkett allegedly helped fast track a visa for the woman's nanny.

Alternative view:

2.54pm: Paul McMullan, the former News of the World deputy, worked on the "name and shame" paedophile story which he says was one of the Rebekah Brooks's good ideas. McMullan offers his definition of public interest: Circulation defines what is the public interest. I don't see it's the job of anyone else to force the public to read this or that. I don't see it's our job to force the public to choose – 'You must read this and you can't read that.'

3.12pm: McMullan says if he hacked David Beckham's phone – which he is not saying he did – it was in the public interest. He also says an ordinary trick used by teenagers has been elevated into a national scandal about phone hacking.