Amplifier Collocations in the Chinese Learner English Corpus

1
Amplifier Collocations in the Chinese Learner
English Corpus

• Jennie, Cai
• 2006-06-02

2
Contents

• Abstract
• I Literature Review
• II Methodology
• III Data Analysis and Discussion
• IV Pedagogical Implication
• V References
• VI Appendix

3
Abstract

• This paper analyzes the usage of amplifiers in
  the Chinese Learner English Corpus (CLEC) by
  analyzing their occurrence in collocations.
• Amplifiers, such as absolutely, completely,
  really, and very, were considered to express
  degrees of increasing intensification upwards
  from an assumed norm
• Maximizers maximally intensify the sense of an
  adjective or verb
• boosters signify less than maximal intensity
  (p.469). (Kennedy,
  203, p. 469)

4
Abstract

• This paper uses the mutual information measure
  (MI) to examine the collocational usage of 23
  amplifiers, including 7 maximizers and 16
  boosters in CLEC.
• maximizers fully, completely, entirely,
  absolutely, totally, perfectly, and utterly
• Boosters very, really, particularly, clearly,
  highly, very much, extremely, badly, heavily,
  deeply, greatly, considerably, severely,
  terribly, enormously, and incredibly.

5
Abstract

• According to Kennedy (2003), these amplifiers are
  the most frequent in the British National Corpus
  (BNC) and all except utterly, severely, terribly,
  enormously, and incredibly occur at least 30
  times per million words (p.472).

6
I Literature Review

• 1. Research Orientation
• Adverbs
• Adverbs of degree
• \
• Amplifiers downtoners
• \
• Maximizers boosters

7
I Literature Review

• 2. Previous Studies
• Klein (1998) claims that Grading of an
  expression is possible if an inherently gradable
  feature is present (p. 9).
• Figure 1
• Table 1

8
I Literature Review

• 2. Previous Studies
• Kennedy (2003, p. 470) introduces recent
  corpus-based researches on amplifier collocations
  by Paradis (1997), Lorenz (1999), Biber et al.
  (1999).
• They begin to notice differing preferences
  across registers and associations with different
  adjectives. (Biber et al., 1999, p. 564, as cited
  by Kennedy, 2003, p. 471.)

9
I Literature Review

• 2. Previous Studies
• Lorenzs (1999) discovers that German learners of
  English may overuse particular modifiers (as
  cited by Kennedy, 2003, p. 470). This inspires me
  to study how Chinese learners of English use
  amplifiers.

10
II Methodology

• 1. Procedure
• In this study, I need the complete file of
  CLEC.txt that contains 1,331,133 running words.
  The corpus tools are Wordsmith 4 and a MI score
  calculator.
• MIlog2 ((f (n, c) N) / (f (n) f (c)).
• N tokens (running words) in text
• F (n) frequency of the keyword
• F (n, c) frequency of the collocation
• F (c) frequency of the word modified

11
II Methodology

• 2. Defection Notes
• Clear (1999) warns that MI score will be very
  unreliable if F (n, c) is low. Low f (n, c) may
  result in high MI score yet, its reliability is
  restricted within the specific corpus.
• For example, decayed only appears twice in CLEC
  and it co-occurs with completely twice, then the
  MI score is as high as 13.88
• Their MI score is less than 3.9 in BNC (Kennedy,
  2003, p. 475).

12
II Methodology

• 2. Defection Notes
• Defection of the Frequency of Verb Collocate

13
III Data Analysis and Discussion

• Absolutely tends to associate with imperative or
  assertive words such as should, necessary, not,
  and no. Only 14.8 collocates are adjectives
  including three positive ones (valuable, good,
  and necessary) and a negative one (negative). Its
  relative frequency in CLEC is only half of that
  in BNC.

14
III Data Analysis and Discussion

• Completely collocates with abolition or negation
  adjectives or participles such as deprived,
  forgotten, destroyed, impossible, and fake, which
  accounts for 40 adjective collocates. The most
  frequent adjective collocateDifferentaccounts
  for 36 adjective collocates. .

15
III Data Analysis and Discussion

• Entirely has only one adjective
  collocatedifferent. The verb CONSIST accounts
  for 22.8 collocates and this is specific to CLEC
  due to the composition topic. Its relative
  frequency in CLEC is less than half of that in
  BNC.

16
III Data Analysis and Discussion

• Fully collocates mostly with verbs. All verb
  collocates are neutral in nature (maybe slightly
  positive, such as agree), which is in contrast
  with Kennedys result that fully has exclusively
  positive associations in BNC. Its relative
  frequency is approximately 1/3 of that in BNC.

17
III Data Analysis and Discussion

• Perfectly has exclusively neutral verb collocates
  whereas it has exclusively positive associations
  in BNC.
• Totally has negative, positive, and neutral
  adjective collocates such as deaf, innocent true,
  and different respectively. This is in contrast
  with Kennedys result that totally tends to have
  mainly negative associations.

18
III Data Analysis and Discussion

• Badly has 65 negative adjective / participle
  collocates (deformed, hurt, polluted, ill) and
  35 neutral ones (insulated, treated, needed).
  The four negative collocates seem to be incorrect
  word choices, yet it reflects the writers
  preference of negative collocates when using
  badly.

19
III Data Analysis and Discussion

• Clearly tends to collocate with perception and
  verbal skills. For instance, remember, know, see,
  hear, understand, and think are all related to
  human perception. Stated and express are related
  to verbal skills. Clearly mainly collocates with
  verbs. Its relative frequency is less than half
  of that in BNC.

20
III Data Analysis and Discussion

• Considerably only has one collocatelong that
  meets the constraints. Its relative frequency is
  just 1/10 of that in BNC.

21
III Data Analysis and Discussion

• Deeply mainly shows its figurative meaning and
  collocates with verbs that indicate influence
  (impressed, affected, encouraged, interested),
  emotion (moved, loved, love, hurt), and
  perception (felt, remember, think, know.) Deeply
  only associate with plough to show its literal
  meaning. Its relative frequency is nearly twice
  of that in BNC.

22
III Data Analysis and Discussion

• Extremely tends to collocate with negative
  adjectives as brutal, expensive, ill, serious,
  and hard. It has positive collocate (important)
  and neutral collocate (round, expensive (negative
  or neutral)). Its relative frequency is less than
  half of that in BNC.

23
III Data Analysis and Discussion

• Greatly tends to collocate with more passive
  verbs and less active verbs. It shows preference
  to positive collocates, yet, it also collocates
  with negative ones (damaged, suffer, decreased)
  and neutral one (changed, different). Its
  relative frequency is about five times of that in
  BNC.

24
III Data Analysis and Discussion

• Heavily has three verb collocates related to
  punishment, strike, and smoke. Its relative
  frequency is a bit more than half of that in BNC.

25
III Data Analysis and Discussion

• Highly mostly collocates with positive words,
  with a neutral adjective (sensitive) as well. Its
  relative frequency is less than half of that in
  BNC.

26
III Data Analysis and Discussion

• Really has mainly positive/neutral collocates,
  with the exception of suffering, sorry, and hard.
  Its relative frequency is nearly 3/4 of that in
  BNC.

27
III Data Analysis and Discussion

• Severely collocates with words related to
  punishment and suffering. The seven occurrence of
  disfigured is specific to one composition topic.
  It seems that severely mainly has negative
  collocates.

28
III Data Analysis and Discussion

• Terribly only has one collocatesorry that meets
  the constraints. Its relative frequency is just
  1/10 of that in BNC.

29
III Data Analysis and Discussion

• Very mainly associates with adjective collocates
  including a lot more positive ones than negative
  ones. Its relative frequency is more than twice
  of that in BNC.

30
III Data Analysis and Discussion

• Very much tends to associate with verb collocates
  that are related to human emotion, such as hate,
  envy, like, love, enjoy, and miss. Another
  frequent collocate is thank because of the fixed
  pattern thank you very much. Its relative
  frequency is more than four times of that in BNC.

31
III Data Analysis and Discussion

• Chinese learners tend to overuse general
  amplifiers such as very, very much, greatly, and
  deeply. On the other hand, learners tend to
  underuse less frequent amplifiers such as
  considerably, fully, and entirely..

32
III Data Analysis and Discussion

• Chinese learners tend to modify verbs with most
  amplifiers, except very, totally, and extremely.
  The lack of adjective collocates may indicate
  that students tend to produce less complicated
  structures, compared with native speakers.
• In addition, students tend to use amplifiers in
  fixed patterns such as terribly sorry, thank you
  very much, completely different, badly needed,
  and think highly. This may indicate that fixed
  patterns have been fabricated in learner
  interlanguage.
• As to the amplifier adjective patterns,
  learners need to be provided with more authentic
  materials, together with the teachers explicit
  awareness-raising instructions.

33
III Data Analysis and Discussion

• Chinese learners tend to associate most
  amplifiers with both negative, neutral and
  positive collocates, ignorant of the
  negative/positive preference of certain
  amplifiers to some extent. For instance, totally
  has mainly negative collocates in BNC yet in
  Chinese Learners English Corpus (CLEC), it
  collocates with positive words like clean, true,
  better, innocent.

34
III Data Analysis and Discussion

• There are word choice mistakes or rare usages
  with adverbs, such as deeply polluted, heavily
  beat, badly polluted and so on. The cause for
  this misuse is obvious once we translate these
  expressions word for word into Chinese. According
  to Gass and Selinker (2001), negative transfer
  causes the misuse (pp. 67-68). Positive transfer
  is apparent as well if we translate the frequent
  collocations such as absolutely necessary,
  clearly stated, deeply moved, greatly helped and
  so on. Language transfer inevitably occur but
  negative transfer is avoidable.

35
IV Pedagogical Implication

• It is not sufficient for students to master the
  usage of modifiers only by memorizing the meaning
  of individual amplifiers and syntactic rules that
  adverbs modify adjectives, adverbs, and verbs. As
  Hunston and Francis (1996) stated that corpus
  studies suggest that all language is patterned
  (p. 14), and Knowles, G. Zuraidah (2004) also
  claimed that relatedness is the starting point
  for the study of the lemma. Thus I suggest that
  pattern usage of amplifiers should also be
  emphasized explicitly through teachers direct
  teaching and implicitly through students
  exposure to authentic materials.

36
IV Pedagogical Implication

• The increasingly larger corpus can serve as a
  huge pool of authentic materials and increasingly
  powerful corpus software can enable us to obtain
  more and more convincing results as EFL
  instructions.
• A comparative study of authentic language data
  and textbooks for teaching English as a foreign
  language has revealed that the use of grammatical
  structures in textbooks differs considerably from
  the use of these structures in authentic English.
  (Dieter Mindt, 1996, 232, as cited by Hoye, 1997,
  231 )
• Therefore, the teacher should always search for
  authentic teaching materials.

37
IV Pedagogical Implication

• In addition, as Firths (1957) famous statement
  that you shall know a word by the company it
  keeps (195). Learners will understand a word
  better by paying attention to its preference to
  negative or positive collocates. As to adverbs of
  degree, Klein (1998) proposes two hypotheses
  regarding this
• (14) adverbs of degree may be specified for
  emotive value if so, they are sensitive to the
  evaluative polarity of the modified element (p.
  3).
• (100) Hypothesis adverbs of degree may be
  specified for emotive value if so, they are
  sensitive to the evaluative polarity of a
  statement. (p. 109).

38
IV Pedagogical Implication

• If a word frequently collocates with negative or
  positive words, then it may acquire a negative or
  positive scent. As a result, it prefers
  negative or positive words even more. Teachers
  and learners should pay attention to the negative
  and positive preference of certain words which
  are not limited to amplifiers. This actually
  belongs to the study of semantic prosody, which
  is a greatly attractive area of research.

39
V References

• Biber, D., Conrad, S. Reppen, R. (2000). Corpus
  Linguistics. Beijing Foreign Language Teaching
  and Research Press.
• Biber, D., Johansson, S., Leech, G., Conrad, S.,
  Finegan, E. (2000). Longman grammar of spoken
  and written English. Beijing Foreign Language
  Teaching and Research Press.
• Bolinger, D. (1972). Degree words. The
  Netherlands Mouton Co., Printers, The Hague.
• Clear, J. (1999). Re Corpora T-score in
  collocational analysis. Retrieved on May, 10,
  2006 from http//torvald.aksis.uib.no/corpora/1999
  -4/0146.html
• Firth, J. R. (1957). Papers in Linguistics.
  London Oxford University Press.
• Gass, S. M. Selinker, L. (2001). Second
  language acquisition. Mahwah, New Jersey, London
  Lawrence Erlbaum Associates, Publishers.
• Hoye, L. (1997). Adverbs and modality in English.
  London and New York Longman.
• Hunston, S. Francis, G. (1996). Pattern
  Grammar. Amsterdam/Philadelphia John Benjamins
  Publishing Company.
• Kennedy, G. (2003). Amplifier Collocations in the
  British National Corpus Implications for English
  Language Teaching. TESOL Quarterly 37 (3)
  467-487.
• Klein, H. (1998). Adverbs of degree in Dutch and
  related languages. Amsterdam / Philadelphia John
  Benjamins Publishing Company.
• Knowles, G. Zuraidah Mohd Don. (2004). The
  notion of a lemma Headwords, roots and lexical
  sets. International Journal of Corpus Linguistics
  9 (1).

40
Thank you very much!

• Questions?