Researchers recently have begun to use the tools and techniques available from corpus linguistics to describe English grammar. This research approach uses computational techniques to study language patterns in a large corpora of texts (see Biber, Conrad, & Reppen, 1998, for an introduction). In most cases, a corpus is designed to represent particular registers (such as conversation, newspaper writing, or academic prose). As a result, grammatical studies based on corpora can describe systematic differences across registers.
Corpus-based studies of English grammar have proven to be especially useful for descriptions of language use, helping us to understand what speakers and writers actually do with the linguistic resources available in English. The findings from corpus-based studies are often surprising. As language teachers and professionals, we often have strong intuitions about language use; for example, we have a good sense of the language features that are typical in conversation, and how speech is different from the language of academic writing. Corpus-based research, however, shows us that our intuitions are often completely wrong. In many cases, we simply do not notice the most typical grammatical features because they are so common.
A related finding from corpus-based research that supports ESP/EAP approaches to teaching is the importance of register for studies of grammar and use. Most functional descriptions of a grammatical feature are not valid for English overall. Rather, strong patterns of use in one register often represent only weak patterns in other registers. From a use perspective, the grammatical features that are typical of conversation are strikingly different from those typical of academic writing, and the same is true for other more specific registers.
In the following sections, I illustrate the interaction of grammar, use, and register with corpus-based analyses adapted from the Longman Grammar of Spoken and Written English (Biber, Johansson, Leech, Conrad, & Finegan, 1999; see also Biber, Conrad, & Leech, 2002). Two case studies are presented, both focusing on the use of verbs: In the next section, one case study deals with lexical patterns (i.e., the most common verbs in English); the section following focuses on the use and distribution of grammatical forms (i.e., the relative frequency of simple, progressive, and perfect aspect in English).
The analyses are based on the Longman Spoken and Written English (LSWE) Corpus, representing four registers: conversation, fiction, newspaper language, and academic prose. Although registers are general, they differ in important ways from one another (e.g., with respect to mode, interactiveness, production circumstances, purpose, and target audience). The corpus used here contains about 20 million words overall, with about 4 to 5 million words from each of these four registers. All frequency counts reported below have been normalized to a common basis—a count per 1 million words of text—so that they are directly comparable across registers.
There are literally dozens of common lexical verbs in English. For example, nearly 400 different verb forms occur over 20 times per million words in the LSWE Corpus (see Biber et al. 1999, pp. 370-371). These verb forms include many everyday verbs such as pull, throw, choose, and fall.
In light of this large inventory of relatively common verbs, it might be easy to assume that no individual verbs stand out as being particularly frequent. However, this is not at all the case: Only 63 lexical verbs occur more than 500 times per million words in a register, and only 12 verbs occur more than 1,000 times per million words in the LSWE Corpus (Biber et al., 1999, pp. 367-378). These 12 most common verbs are say, get, go, know, think, see, make, come, take, want, give, and mean.
Figure 1 indicates the importance of these 12 verbs by plotting their combined frequency compared with the overall frequency of all other verbs. Taken as a group, these 12 verbs are especially important in conversation, where they account for almost 45 percent of the occurrences of all lexical verbs. Obviously, any conversational primer that did not include extensive practice of these words would be shortchanging students.
There are also large frequency differences among these 12 verbs, overall and in their register distributions. See Figures 2 and 3, which plot the frequency of each verb in conversation and in newspaper language (cf. Biber et al., 1999, pp. 374-376).
The verb say is listed first in these figures because it is common in both spoken and written registers and thus has the highest frequency overall, which is not surprising, given the ubiquitous need to report the speech of others. Both speakers and writers rely heavily on the single verb say for this purpose, usually in the past tense expressing either a direct or indirect quote. For example:
You said you didn’t have it. (Conversation)
He said this campaign raised “doubts about the authenticity of free choice.” (News)
The extremely high frequency of the verb get in conversation is more surprising. This verb goes largely unnoticed, yet in conversation it is by far the single most common lexical verb. The main reason that get is so common is that it is extremely versatile, with a wide range of meanings. For example:
See if they can get some of that beer. (Conv)
They’ve got a big house. (Conv)
Get in the car. (Conv)
So I’m getting that way now. (Conv)
Several other verbs are also extremely common in conversation: go, know, and, to a lesser extent, think, see, come, want, and mean. News, on the other hand, shows a quite different pattern, with only the verb say being extremely frequent. However, it should be noted that all 12 of these verbs are notably common in both registers in comparison to most verbs in English. For example, as noted above, verbs such as pull, throw, choose, and fall occur only about 50 to 100 times per million words. Countless other verbs have even lower frequencies. In contrast, the majority of the 12 most common verbs occur over 1,000 times per million words in both conversation and news.
Thus there is a cline in the use of verbs: A few verbs occur with extremely high frequencies; several verbs occur with moderately high frequencies; and most verbs occur with relatively low frequencies. In addition, different registers show strikingly different preferences for particular verbs. For example, the verbs get, go, know, and think are much more frequent in conversation than in news (see Figures 2 and 3 above). In contrast, verbs such as add, spend, claim, and continue are much more common in news than in conversation.
One of the most widely held intuitions about language use among English-language professionals is the belief that progressive aspect (e.g., the present continuous) is the unmarked choice in conversation. This belief is sometimes reflected in the overly frequent use of progressive verbs in made-up dialogues (like those found in older ESL/EFL coursebooks teaching conversation skills). For example,
Doctor: Hello Mrs. Thomas. What can I do for you?
Patient: Well, I’ve been having bad stomach pains lately, doctor.
Doctor: Oh, I’m sorry to hear that. How long have you been having them?
Patient: Just in the past few weeks. I get a very sharp pain about an hour after I’ve eaten.
Doctor: Well, I don’t think it’s anything serious. Maybe you eat too quickly. You don’t give yourself time to digest your food.
Patient: My husband is always telling me that.
As Figure 4 shows, progressive aspect is in fact more common in conversation than in most written registers. The contrast with academic prose is especially noteworthy: Progressive aspect is extremely rare in academic prose but more common in conversation. However, the use in other written registers is somewhat surprising: Progressive verb phrases are nearly as common in fiction as in conversation, and they are relatively common in news as well.
We might wonder, though, about the use of progressive verbs in relation to simple aspect verbs. Many English teachers (including myself!) have had the intuition that progressive aspect verbs are the normal choice in conversation, and thus they should be more common than simple aspect verbs in that register. In this case, corpus research shows that our intuitions are dramatically wrong: As Figure 5 shows, simple aspect verb phrases are more than 20 times as common as progressives in conversation.
The following excerpt from the LSWE corpus illustrates the normal reliance on simple aspect in natural conversation:
B: Yeah. Do you -
A: No I’m, I’m only on the PTA.
A: That’s it.
A: Yeah.
In contrast, progressive aspect is used for special effects, and usually focuses on the fact that an event is in progress or about to take place. For example:
But she’s coming back tomorrow. (Conv)
With nondynamic verbs, the progressive can refer to a temporary state that exists over a period of time, as in the following:
I was looking at that one just now. (Conversation)
You should be wondering why. (Conversation)
We were waiting for the train. (Conversation)
A few lexical verbs—such as bleeding, chasing, shopping, starving, joking, kidding, and moaning—actually occur most of the time with the progressive aspect in conversation. However, the norm—even in conversation—is to express verbs with the simple aspect. In marked contrast to the expectations created by some popular conversational materials, verb phrases such as I’ve been having and is always telling are exceptions rather than rules.
Awareness of the patterns of language use revealed by corpus research is obviously important for both teachers and students. However, frequency information cannot necessarily be mechanically translated into materials for instruction and assessment. Other considerations, such as the ease/difficulty of learning, are also important. At the same time, we can also no longer afford to ignore the typical patterns of use identified by quantitative corpus analysis. Instead, we can look forward to important gains for students as we begin to develop materials that reflect the actual patterns of use in particular registers.
References
Biber, D., Conrad, S., & Reppen, R. (1998). Corpus linguistics: Investigating language structure and use. Cambridge: Cambridge University Press.
Biber, D., Conrad, S., and Leech, G. (2002). The Longman student grammar of spoken and written English. London: Longman.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spoken and written English. London: Longman.
Richards, J. & Hull, J. (1997). As I was saying ... Conversation tactics. New York: Addison-Wesley.
Marianne Celce-Murcia, University of California, Los Angeles, and American University of Armenia, celcemurc@aol.com
The study of grammar (i.e., morphology and syntax) in formal linguistics has been based on the assumption that grammatical systems are innate, autonomous, and context-free (Chomsky, 1965, 1968). One result of this perspective has been that grammatical analysis has been restricted to the sentence level. For purposes of teaching communicative use of a foreign or second language, it is necessary to go beyond the sentence level, as listeners, speakers, readers, and writers possessing communicative competence do not restrict themselves to the sentence level when they produce and interpret messages.
I agree with functionalists, who claim that lexico-grammar is an integral part of interpreting and producing discourse in context (Halliday, 1994). In fact, I have argued elsewhere (Celce-Murcia, 1991) that the majority of rules for grammatical forms—in English or any language—are context-sensitive2 and that we need to do careful qualitative analyses of grammar in discourse to identify these contexts.
choice of logical connector
My focus here comes under the first item on the above list (i.e., tense-aspect-modality choice). The present perfect progressive (PPP) tense-aspect combination is a fairly complex form in this category. It is highly context-sensitive and because it does not occur with high general frequency, it is of very marginal interest in corpus linguistics.4I will show that it is nonetheless a crucial form for learners of English who wish to acquire conversational competence.
If we have a sentence with the PPP that occurs out of context, it can be very difficult to interpret:
John has been walking.
What does this sentence mean? John has been walking for 1 hour and is still walking? John has been walking lately (to get some exercise)? John has been walking (around the room to rid himself of a leg cramp)? If we put this sentence into an unfolding discourse, which is one common way of providing context, the sentence becomes interpretable:
A: What has John done to lose so much weight?
B: He’s been walking.
The context tells us that this activity has been repeated regularly by John over a period of time but that the activity is not necessarily in progress at the moment of speech. Much of the English tense-aspect-modality system is similarly context-dependent in terms of any meanings or uses that one can assign (see Celce-Murcia & Larsen-Freeman, 1999, pp. 122, 161-181).
Although I am focusing specifically on the PPP tense-aspect form here, I also mention other less complex tense-aspect forms as part of the conventionalized discourse sequences I discuss in this paper.
Background on Verb-Phrase Semantics
In this qualitative analysis and explanation of how the PPP is used in English, I will sometimes refer to the type of verb phrase (VP) the PPP co-occurs with. Thus I digress briefly as it is important to determine whether the base form of the verb phrase occurring with PPP expresses a state, a punctual action, an accomplishment, or an activity (Vendler, 1967), as illustrated in the following examples:
VP of state: Ralph understands the problems.
VP of punctual action: Mary broke the cup.
VP of accomplishment: Susan baked a cake.
VP of activity: John walks.
Of these four VP types, VPs of state and VPs of punctual action occur least frequently with the PPP. When the PPP does occur with such VPs, their semantic classification gets modified:
VP of state becomes change of state over time:
Ralph has been understanding the problems (better lately).
VP of punctual action becomes reiterated action over time:
Mary has been breaking the cups (ever since she started waitressing here).
VPs of accomplishment occur readily with the PPP; however, the use of this tense-aspect form also modifies the semantics of the VP by signaling that the outcome or product is either not yet completed or has just very recently been completed:
Susan has been baking a cake.
(Either the cake is still baking in the oven or the speaker, who can smell the aroma of a cake baking, says the above sentence, and then walks into the kitchen and sees the cake cooling on the counter.)
VPs of activity are the most frequent and natural VP type that occurs with the PPP:
John has been walking.
In such cases the PPP does not change the semantics of the VP; however, we need to know the situational context or the discourse context to be able to give the sentence the proper interpretation, which is what I did with this sentence earlier when I contextualized it by first asking what John had done to lose weight. In fact, activity VPs predominate in the discourse sample I present below to contextualize the PPP and to teach it.
Describing and Teaching the Use of PPP
In light of the highly contextualized nature of the verb phrases occurring with PPP, it is important that we examine a large number of naturalistic discourse segments in which the PPP occurs. This is the only way we can hope to arrive at an understanding of how and when native English speakers use this structure. For a full understanding we must analyze more than the sentence in which the form occurs. We must analyze the situational context and the entire discourse episode in which the form occurs so we can see what precedes it and what follows.
This need to analyze the PPP within larger discourse episodes also means that we cannot teach the PPP using sentence-level exercises. I thus recommend that English language teachers make use of authentic texts that contain tokens of the PPP from sources such as transcribed speech, comic strips, letters, advertisements, fiction, and drama. However, it is wise to keep in mind Ur’s (1984) distinction between “genuine authentic” materials and “imitation authentic” materials as some minor editing or simplification often makes such materials more accessible and much more effective pedagogically if one wants to make a grammatical form particularly salient to learners.
The PPP is a somewhat complex structure that is not presented to beginners. In fact, intermediate-level learners need mastery of the simple present tense, the progressive aspect (be + -ing), and the perfect aspect (have + past participle) before they are ready to take on the PPP. Once they are ready, their learning task will be much easier if they are presented with realistic everyday examples of the PPP that they can understand and discuss before they are expected to use the target form in similar situations.
I’ll offer one discourse-based context with activity types for teaching and learning the PPP, drawn from some of the naturalistic discourse segments analyzed in the study reported in Celce-Murcia and Yoshida (2003), which was based on 250 contextualized tokens of PPP.
Example: Talking About Activities
Consider the following transcribed segment from a radio talk-show in Los Angeles,5 which I have used to teach this discourse function of the PPP:
Host: I’m Dr. Mary Smith and you’re on talk radio. Hello?
Caller: Good afternoon, Dr. Smith. I’m Lucy and I’m 59. I’d just like to share a positive thing that I have found in the last year or so. I’ve been hiking and find that this is a wonderful way to keep your weight down and meet some people and just really feel good. (cited in Celce-Murcia & Yoshida, 2003, p. 6)
First my students listen to the recording several times and retell in their own words what the interlocutors are talking about. They ask questions about any vocabulary or grammar they find problematic. Then I show them the transcript and tell them that Lucy’s segment follows a typical pattern for use of the PPP:
-First move: present context + past information (present or present perfect tense)
-Second move: activity of focus, speaker’s special concern (PPP)
-Third move: evaluation/follow-up (present tense)
Then the students identify and label each sentence in Lucy’s transcript according to which move it belongs to (e.g., the first two sentences after the greeting are the first move).
To prepare for the students’ own use of the PPP, we talk about the importance of physical activity for good health; then they work in groups to identify an activity they have been doing recently that they can relate to their group members much as Lucy did in the radio talk-show. My students have come up with activities such as the following:
-I’ve been jogging.
-I’ve been swimming every day.
-I’ve been playing tennis with a friend.
-I’ve been riding my bike to school.
Their discourse-level assignment is to develop a three-move dialogue in which the opening question provides the context or the background (i.e., the first move):
What have you been doing lately (to keep fit)?
One short but completely appropriate response to this question was:
I’ve been jogging, and it helps keep me fit.
The first clause of the response is the focus activity (i.e., the second move) and the second clause is an evaluation or follow-up (i.e., the third move).
In fact, the three-part structure that I have just outlined for talking about activities is so salient that under some circumstances, one can utter simply the focus activity (i.e., the second move) with the first move (background) and the third move (evaluation) being implied. For example:
Move 1 (background, i.e., shared information)
Move 2 “You’ve been smoking again.”
Move 3 (implied but unspoken negative evaluation)
This is just one possible example. For similar descriptions and teaching suggestions for two other salient discourse-based uses of the PPP (i.e., complaints and explanations), see Celce-Murcia and Yoshida (2003).
Conclusion
The example and activities that I have suggested for teaching the PPP are very different from what is normally found in most ESL/EFL textbooks. The description and teaching activities that I have discussed for presentation, analysis, and production of this form have all been embedded in context and have come from naturalistic discourse. To be effective, grammar instruction needs to simulate what actually occurs in spoken and written communication. Learning activities should help students understand how grammar is a resource for shaping purposeful discourse. To achieve the goal of communicative competence in a second or foreign language, students’ learning activities need to be pragmatically well motivated in terms of the social contexts and the language forms used; only a careful qualitative analysis can identify the most typical contexts—that is, the ones to use in teaching. This is the direction that language teachers and materials writers need to take in order to improve grammar instruction, not only of the PPP but of most other structures typically treated in an ESL/EFL curriculum. The discourse level is where learners can best process, understand, apply, and ultimately internalize the ways in which grammar is used in their second or foreign language. (See also Celce-Murcia & Olshtain, 2000, p. 68.)
Endnotes
1. I would like to acknowledge that this paper draws very heavily on prior research that I carried out with Nina Yoshida (Celce-Murcia & Yoshida, 2003). I am grateful to Nina for her collaboration but take full responsibility for all errors in this paper.
2. Very few grammatical rules are strictly local and sentence-internal in English. Rules such as determiner-noun agreement, use of reflexive pronouns as objects, and use of gerunds after prepositions are examples of the exceptions to my generalization. All languages seem to have a small number of such local agreement rules, with the majority of grammatical rules being context-sensitive.
3. Some people think subject-verb agreement is strictly a sentence-level rule in English; however, certain types of noun subjects can be construed as either singular units or as plural collectives, causing variability in this agreement rule:
-The gang is causing riots.
-The gang are causing riots.
Thus, because this rule (which is often as challenging for native speaker as it is for nonnative speakers) frequently depends on the speaker or writer’s perception of the number of the subject noun, it is not, strictly speaking, a mechanical agreement rule.
4. See Biber, Johansson, Leech, Conrad, and Finegan (1999), where only two brief mentions of the PPP come up in this 1,204-page long corpus-based grammar of English:
p. 462: “Perfect progressive verb phrases (marked for both aspects, e.g., has been waiting) are rare in all registers, comprising less than 0.5% of all verb phrases.”
pp. 464-5: “Additionally have/has been functions as the auxiliary part of a passive verb phrase or (less commonly) of a progressive verb phrase:
However once its position has been determined it is in a totally different state. (ACAD)
Hunter Howeston, an American secret service agent, has been reminiscing about the days when counterfeiting was a specialist art form. (NEWS)”
5. All proper names in our naturalistic oral data have been changed to protect the identities of the speakers.
References
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spoken and written English. London: Longman.
Celce-Murcia, M. (1991). Discourse analysis and grammar instruction. Annual Review of Applied Linguistics, 11, 135-151.
Celce-Murcia, M., & Larsen-Freeman, D. (1999). The grammar book: An ESL/EFL teacher’s course (2nd ed.). Boston: Heinle & Heinle.
Celce-Murcia, M., & Olshtain, E. (2000). Discourse and context in language teaching. New York: Cambridge University Press.
Celce-Murcia, M., & Yoshida, N. (2003). Alternatives to current pedagogy for teaching the present perfect progressive. English Teaching Forum 41(1), 2-9, 21.
Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.
Chomsky, N. (1968). Language and mind. New York: Harcourt, Brace, and World.
Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward Arnold.
Ur, P. (1984). Listening and language learning. Cambridge: Cambridge University Press.
Vendler, Z. (1967). Linguistics and philosophy. Ithaca, NY: Cornell University
Grammar: Rules or Patterning? Norbert Schmitt, University of Nottingham, Norbert.Schmitt@nottingham.ac.uk
What is grammar? Grammar has been traditionally thought of as a large number of rules that describe the syntax of a language, typically at the sentence level. Indeed, virtually every grammar book describes such rules. However, I like to think of grammar in broader terms. Instead of limiting myself to the traditional realm of grammar, I like to think of the total systematicity of language, that is, everything that is systematic and regular about language. Of course, there must be regularity in language; otherwise every person would be constructing it in different ways, leading to a total breakdown in communication!
Language systematicity extends from the smallest components all the way up to extended discourse. But when we look at the various levels of language, the systematicity we find often does not conform to absolute rules. There always seem to be exceptions that spoil any attempt at a rule-based description. (It is probably better to think in terms of probability-based regularities, instead of fixed rules, but even regularities do not always adequately describe the behavior of language.) There is also an element of widespread patterning, which often accounts for the systematicity of language better than rules do.
Let us look at the various levels of language from a patterning perspective. In writing, the smallest unit is letters, and they combine to make up syllables and words in certain ways. Below is a list of six words, some of which I made up and some of which are real, but very low frequency. Can you tell which is which?
prolificity hgough
tcharal nulliparous
quintain louqt
Although it is unlikely that you were familiar with any of these word forms, your intuition probably suggested that tcharal, hgough, and louqt are not real words. The thing that probably seemed strange about these word forms is the unusual consonant clusters: tch and hg are not productive clusters at the beginning of English words, and qtdoes not normally end words. These are not fixed spelling rules (someone could invent something and name it any of these word forms), and you almost certainly did not learn from any textbook that these clusters are "incorrect." Still, when I do this activity at presentations, the audience is always able to spot the word forms that do not contain the normal patterning of English spelling. Thus, even at this most basic level, it seems clear that there is systematicity in language, and that it is governed at least to some extent by patterning rather than rules.
We find similar patterns at the level of morphology, where suffixation often does not conform to any firm rule. Of course there are some fairly regular transformations. For example, changing verbs ending in -ate (punctuate, elevate) into nouns normally results in -ion forms (punctuation, elevation). However, many transformations fall into a number of patterns rather than come under any single rule. For instance, the noun form of adjectives ending in -al can end in -ness (casual > casualness) or -ity(formal > formality), or be in the root form (influential > influence).
At more global levels of language, patterning becomes even more significant and also harder to identify solely by intuition. Let us examine the word border to illustrate this. If your students ask what border means, you would probably say something like "the edge or boundary of something." You might also show them various inflections of the word (bordered, bordering, borders). Because infections are usually considered a grammatical change, you might assume that they all have a similar meaning in context. But you would be wrong.
If we look at the behavior of the border word family in the British National Corpus (BNC; a 100-million-word corpus of English), we come up with the following figures:
| BNC frequency | X + "on" | Figurative sense |
| border | 8,011 | 89 (1%) | |
| borders | 2,539 | 84 (3%) | |
| bordering | 367 | 177(48%) | 71% |
| bordered | 356 | 99 (28%) | 75% |
From these figures we can see that border and borders (mainly noun forms) are the most frequent members of the family, which is not at all surprising as most word families have more and less frequent members. However, once we put the words into phrases (in this case by adding the preposition on), the behavior changes dramatically. Only 1 to 3% of the cases of border and borders occur in combination with on, but about one quarter of the occurrences of bordered do, as do almost one half of the occurrences of bordering. Clearly there is a strong tendency for bordered and bordering to occur in a pattern with on. But the patterning not only involves the combination of the words; it also affects the meaning. Whereas border and borders almost always refer to the expected or literal meaning of "edge" or "boundary" (even when in combination with on), in about three quarters of the cases, bordering on and bordered on refer to some figurative meaning not to do with edges or boundaries. In fact, when we look at concordance lines from the BNC, we find quite a different usage:
- His passion for self-improvement bordered on the pathological.
- But his approach is unconscionable, bordering on criminal.
For further evidence of this usage, here are some other words that occur to the right of bordered/ing on:
a slump arrogance chaos
a sulk austerity conspiracy
acute alcoholic poisoning bad taste contempt
antagonism blackmail cruelty
apathy carelessness cynicism
There is clearly a pattern here, and I would suggest that it is something like this:
SOMETHING/ (be) bordered/bordering on AN UNDESIRABLE STATE
SOMEONE (OFTEN OF MIND)
The main point of these examples is that there is a systematicity around the use of bordered and bordering that is not captured by a traditional grammatical description. The structure (noun phrase + BE + bordered/bordering + preposition + noun phrase) does not really tell us much about the way these words are used. In contrast, the above pattern-based description tells us much more about how the words are used in context and what they mean.
I do not want to give the impression that rules are not an important part of grammar, or that the mind does not have access to some kind of rule-based system. However, I hope that I have shown that the systematicity of language is not solely rule-based, and that a large degree of patterning is present that is not described well by rules. In fact, the more we look at corpus evidence, the more patterning we find. We may discover in the end that patterning actually makes up the majority of the systematicity of language, with rules coming into play only when insufficient patterning is available. But whatever the case, patterning is a key component of language, and any view of grammar needs to take this into account.
Acknowledgment: John Sinclair did the original analysis of "border" upon which this expanded version is based.
Research and Theoretical Perspectives on Output Practice in Grammar Teaching Diane Larsen-Freeman, University of Michigan, dianelf@umich.edu
Grammatical practice has not received much attention from applied linguistic researchers; however, I think it should. Current SLA research perspectives tend to favor input-processing rather than skill-building as a means of developing grammatical competence. Though I would not wish to challenge the contribution of input-processing, I believe that output practice deserves to be investigated as well. Output practice was central to the Audiolingual Method, and although I do not recommend a restoration or perpetuation of meaningless, soporific drills that accompanied the view of language as verbal behavior, the fact that output practice continues to be a well accepted teaching practice in a postbehaviorist world justifies its investigation by applied linguists.
It has long been held in language teaching, and most learners would attest to this, that comprehension usually precedes production; however, many researchers act today as if comprehension or consciousness-raising guarantees accurate production. Clearly, this is not the case. Of course, this is not to say that there is no overlap between the two, but there is evidence that input processing and speech production require distinct types of processing mechanisms (White, 1991). Other research suggests that comprehension skills and production skills are to some extent learned separately (DeKeyser & Sokalski, 2001). And yet, as I say, output practice has virtually been ignored or dismissed by SLA researchers, who might say that practice promotes fluency, but does not play any substantive role in the development of the underlying grammatical system. Though it is worth considering what arguments have been made for and against practice by researchers, I do not have the space here to go into any detail. Let me, therefore, summarize in tabular form the benefits and drawbacks of output production/practice that I have culled from selected research literature up to the year 2000 (for a fuller explication and other references, see Larsen-Freeman, 2003).
The Benefits and Drawbacks to Output Production/Practice
Second Language Acquisition Research
(adapted from Larsen-Freeman, 2003)
| Benefits | Drawbacks |
| Moves learners from semantic to syntactic processing; encourages syntactic analysis (Swain, 1985; Swain & Lapkin, 1995) | Is not needed for language acquisition, or at best is out of sync with the natural development of grammatical competence (Krashen, 1982; R. Ellis, 1993, 1998) |
| Promotes noticing (Long, 1996; Schmidt, 1990) | Is scarce in the classroom (Krashen, 1998) |
| Used by learners to test hypotheses and gain negative feedback through which to modify their hypotheses (Schachter, 1984) | Pushes learners to speak before they are ready, which might lead to negative affect and misrepresentation of the grammatical rule (Krashen, 1982, and Eubank, 1987, as cited in R. Ellis, 1998) |
| May help learners gain more comprehensible input or better access to the developing system (Krashen, 1982) | Does not directly affect the system itself (White, 1991, in Braidi, 1999, p. 135; Long, 1996; Schwartz, 1993; VanPatten & Cadierno, 1993) |
Notice that in the SLA research literature even the benefits present a rather weak endorsement of output production (and not much at all of output practice), suggesting that it leads only indirectly to a modification of underlying competence through promoting awareness and analysis and allowing for more input and hypothesis-testing, but not in itself directly altering competence. However, stronger theoretical justification for output practice can be found in cognitive psychology, particularly in research from an information-processing perspective.
The Benefits and Drawbacks to Output Production/Practice
Cognitive Psychology (Information-Processing Perspective)
| Benefits | Drawbacks |
| Helps learners develop fluency through the control of formulaic speech (Gatbonton & Segalowitz, 1988) | |
| Increases automaticity, which in turn frees up attentional resources (Anderson, 1985) | |
| Leads to restructuring, which modifies and reorganizes underlying representations (McLaughlin, 1987, 1990) | Can result in a (hopefully temporary) degradation in performance (McLaughlin, 1987) |
As to the last benefit in the chart above, McLaughlin (1987, p. 136) wrote:
But there is more to learning a complex cognitive skill than automatizing subskills. The learner needs to impose organization and to structure the information that has been acquired. As more learning [in 1990, he says “practice”] occurs, internalized, cognitive representations change and are restructured. This restructuring process involves operations that are different from, but complementary to, those involved in gaining automaticity.
Thus, from an information-processing perspective, practice can indeed alter the underlying system.
Another theory originating in cognitive psychology, connectionism, contributes to the list of benefits as well, hypothesizing that practice strengthens connections in a neural network.
The Benefits and Drawbacks to Output Production/Practice
Cognitive Psychology (Connectionist Perspective)
| Benefits | Drawbacks |
| Leads to repeated neural network activation, which, in turn, results in stronger, and therefore more easily activated, connections (N. Ellis, 1996) | |
The final theoretical justification I offer here stems from a Chaos/Complexity Theory (C/CT) or Dynamic Systems Theory perspective. Much in these theories is of interest to applied linguists (see, for example, Larsen-Freeman & Cameron, forthcoming). For my purposes here, though, I would point out that both information-processing and connectionist perspectives stop short of a more radical position that I have been drawn to for almost a decade (see, for example, Larsen-Freeman, 1997). From a C/CT perspective, one might argue that the language system not only is restructured or reweighted as a result of use, but is created. For in a dynamic system, it is not just that the state of a system changes over time; so does the relationship among the elements that constitute it, such as happens with a developing embryo. After all, language is not a closed, entropic system. It does not settle down to a point of equilibrium. Instead, as with other naturally occurring systems, language is dynamic, constantly evolving, emergent, and self-organizing. “We create language as we go, both as individuals and as communities” (Harris in Bybee & Hopper, 2001, 19). Thus, every use of language by language learners or language users changes its resources.
As Mohanan (1992, pp. 653-654) put it,
Suppose we free ourselves from the idea that [first] language development is the deduction of the adult grammar from the input data, and think of it as the formation of patterns triggered by the data.
In other words, rather than grammar development being solely a process of conforming to the grammar of the community, which is governed by deductive (and inductive) operations, language use/development involves the creation of grammatical patterns (morphogenesis), which then, as speakers communicate with each other, adapt themselves to the overt patterns of the grammars of other individuals in the community.
Therefore, practice is not exclusively for the purpose of rehearsing mental competence to promote its fluency or its restructuring. Practice can create new resources that learners can call upon to foster their participation.
The Benefits and Drawbacks to Output Production/Practice
Chaos/Complexity Theory
| Benefits | Drawbacks |
| Results in morphogenesis, the creation of new patterns (Larsen-Freeman, 1997; Mohanan, 1992) | |
In conclusion, although research on output practice in grammatical instruction is increasing these days, there is still much work to be done in order to investigate the theoretical claims summarized in this article, which are as follows: from an information-processing perspective, practice activities are essential in language teaching because they encourage automaticity, which frees learners’ attention to be directed elsewhere, and may contribute to restructuring learners’ grammars. From the perspective of connectionism, practice strengthens the connections among the nodes in a neural network, accelerating future access. From the perspective of C/CT, practice may lead to the creation of new language patterns. All these benefits are not likely to accrue if practice is meaningless and soporific, of course, but only if the practice is meaningful and engaging. What this takes, though, is a subject for another paper.
References
Anderson, J. (1985). Cognitive psychology and its implications (2nd ed.). New York: W.H. Freeman and Company.
Braidi, S. (1999). The acquisition of second-language syntax. London: Arnold.
Bybee, J., & Hopper, P. (Eds.) (2001). Frequency and the emergence of linguistic structure. Amsterdam/Philadelphia: John Benjamins Publishing Company.
DeKeyser, R., & Sokalski, K. (2001). The differential role of comprehension and production practice. Language Learning, 51(Suppl. 1), 81-112.
Ellis, N. (1996). Sequencing in SLA: Phonological memory, chunking, and points of order. Studies in Second Language Acquisition, 18(1), 91-126.
Ellis, R. (1993). Second language acquisition and the structural syllabus. TESOL Quarterly, 27(1), 91-113.
Ellis, R. (1998). Teaching and research: Options in grammar teaching. TESOL Quarterly, 32(1), 39-60.
Gatbonton, E., & Segalowitz, N. (1988). Creative automatization: Principles for promoting fluency within a communicative framework. TESOL Quarterly, 22(3): 473-492.
Krashen, S. (1982). Principles and practice in second language acquisition. Oxford, UK: Pergamon.
Krashen, S. (1998). Comprehensible output? System, 26, 175-182.
Larsen-Freeman, D. (1997). Chaos/Complexity science and second language acquisition. Applied Linguistics, 18(2), 141-165.
Larsen-Freeman, D. (2003). Teaching language: From grammar to grammaring. Boston, MA: Heinle/Thomson.
Larsen-Freeman, D., & Cameron, L. (forthcoming). Applied linguistics and chaos/complexity theory. Oxford University Press.
Long, M. (1996). The role of the linguistic environment in second language acquisition. In W. Ritchie and T. Bhatia (eds.), Handbook of second language acquisition (pp. 413-468). San Diego: Academic Press.
McLaughlin, B. (1987). Theories of second-language learning. London: Edward Arnold.
McLaughlin, B. (1990). Restructuring. Applied Linguistics, 11(2), 113-128.
Mohanan, K. P. (1992). Emergence of complexity in phonological development. In C. Ferguson, L. Menn, & C. Stoel-Gammon (Eds.), Phonological development (pp. 635-662). Timonium, MD: York Press.
Schachter, J. (1984). A universal input condition. In W. Rutherford (Ed.), Universals and second language acquisition. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11(2), 129-158.
Schwartz, B. (1993). On explicit and negative data effecting and affecting competence and linguistic behavior. Studies in Second Language Acquisition, 15(2), 147-163.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in development. In S. Gass and C. Madden (Eds.), Input in second language acquisition (pp. 235-253). Rowley, MA: Newbury House.
Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied Linguistics, 16(3), 371-391.
VanPatten, B., & Cadierno, T. (1993). Input processing and second language acquisition: A role for instruction. Modern Language Journal, 77, 45-57.
White, L. (1991). Adverb placement in second language acquisition: Some effects of positive and negative evidence in the classroom. Second Language Research, 7,133-161.
Community News and Information Open Leadership Positions The Applied Linguistics Interest Section (ALIS) would like to invite volunteers to run for the following positions: chair-elect and newsletter coeditor. The elections will be held at the business meeting at the 40th Annual TESOL Convention in Tampa, FL, in the United States. Candidates should expect to attend the business meeting. A 3-year commitment is expected for both positions (or possibly 2 years, in the case of newsletter coeditors). If you are interested, have some expertise or experience, and have time to devote to TESOL, please contact Past Chair David Olsher (olsher@sfsu.edu) or Chair Noël Houck (nrhouck@csupomona.edu).
Chair-Elect
The chair-elect serves for 3 years, first as chair-elect, then as chair, and then as past chair, and has different duties with each job. The chair-elect is expected to attend leadership meetings at the annual TESOL convention the year he or she is elected and again the following 2 years. The main duties as chair-elect are organizing the academic session panel for the annual TESOL convention and, in cooperation with the chair, organizing the discussion sessions. This task involves inviting a variety of discussion leaders to supplement submitted proposals. In his or her second year, the chair is responsible for managing the review process for submissions for the ALIS concurrent sessions, leading the open business meeting, participating in the IS Council meeting, and working with the chair-elect and newsletter editors as needed to plan conference events (InterSections, special projects) and communicate with ALIS members. Finally, the past chair acts as consultant and advisor to the new chair and chair-elect.
Newsletter Coeditor (one position)
The two newsletter coeditors, serving for 3 years each, approach applied linguists and researchers and solicit from them articles of interest for the newsletter. They then compile, edit, and proofread the articles and finally submit them to TESOL Central Office for posting to the online newsletter. Coeditors should expect to attend every annual ALIS business meeting and TESOL editorial workshop (at the annual TESOL convention) during the time of their appointment. Highly desirable skills for this job include an eye for detail and staying in consistent correspondence with each other, article authors, chairs, and TESOL Central Office.