Thomas Leverett, 2008

[ home ][ writing ][ weblog ]

A number of different language learners and their struggles have stayed with me; their quest to become fluent in English has been a rough and painful road. The first is below; a Chinese speaker has trouble with relative clauses, and also with the construction It is difficult (for me) to do this. By trouble I don't mean he had trouble understanding them; I have no evidence of that, not having been there if they were ever explained. I mean that, at a high level, this learner should be making these easily, and is not. As with many of these problems, a traditional language teacher would point to interference as the root of the problem: the Chinese language has other ways to do this, and those ways get in the way, when this speaker/writer tries to make sentences.

Second language educators used to talk about interference as a driving force behind most second language errors. The first language had simply gotten in the way; assumptions that a person brought forward from his or her first language held until he/she picked up the appropriate habits, and in the meantime his/her second language looked a lot like the first, only with second-language translations of first-language, often inappropriate, words.

The concept is quite useful in that we can see entire populations of first-language Chinese, for example, make the same errors with English; these errors are predictable, and happen with virtually every English learner in any given group, but there is an interesting twist; some people break some habits before others. For Chinese, for example, it is difficult to master relative clauses (I know the man who(m) she met) as Chinese doesn't use them to express the same ideas (instead, he might say: I know the man she met him…and, as another example, Chinese tend to say I am difficult to study English and must unlearn this and learn how to say It is difficult for me to study English and similar patterns.

If there is indeed a uniform order of acquisition, such that all learners learn all grammatical functions in the same order, presumably most complicated ones last, then these would all be picked up in predictable places in Chinese students' learning path, and we would be able to predict when each first-language Chinese speaker was ready to pick up a grammatical form. This, however, is not the case. Virtually all Chinese have trouble with the above constructions, but they pick up the right forms in their own time, in different orders, in their own time. The principle here can be stated simply as:

Any two language learners, given the same native language and similar learning environments, may still acquire grammatical structures in different orders from each other.

I make this claim after watching language learners carefully for many years, but still realize that there are many other outside influences. Below are a few.

First, each may have more than one native language; this is frequently the case with native Chinese speakers, but is also true of a host of other language learners. I don't know all of the various dialects of Chinese, including Taiwanese, Cantonese, etc., but have taught in Korea and other more monolingual environments where entire classrooms of learners shared one and only one language.

Second, learning environments, though similar, are not exactly the same; the Chinese above, for example, may be in a situation where they have to say that something is difficult or easy, hundreds of times a day; this alone may drive their acquisition of a single structure, and make it earlier than it would be under more normal conditions. Similarly, one person may have a teacher who drives a point home in such a salient way as to influence a student's acquisition (I always hope this, as a teacher, yet, from my own experience, know that it's more likely to be a real-life situation than a classroom situation that is likely to burn a new grammatical structure into my consciousness).

My main point is that, even in a large population of monolingual second-language learners, there are enough variables to affect each and every learner differently, and make each path different in turn.

A closer look at the structures above point out what to me are interesting and common problems in language learning. Krashen would hypothesize that one of the structures would be acquired before the other, as a matter of natural course; this order would presumably be based upon inherent complexity of a structure, or simply the fact that languages spring up from nowhere, like a card castle, and one must have the lower-order things in place before mastering the higher-order ones. But this general conception of language breaks down fairly quickly when one looks at structures like the two above. In what way is one more complex than the other? Would acquisition of one of the structures be necessary, before acquisition of the other could take place? Here I have a couple of working principles which I believe challenge the established order.

Acquisition of some grammatical structures may require mastery of more basic sentence structures, but language is actually not that complex; most language structures can be acquired and mastered with little or no background understanding.

Rather than adopting a conception of language and its acqusition as a card-castle with nine layers, all of the top seven depending on the stability of a single card on the second, and another on the fourth, a better conception is that of a wide, shallow castle with many simple structures relying on a single stable first floor, that basic outline being simple sentence order, or simple placement of most structures.

Acquisition of a given structure, for example one of the above, does not occur at a fixed time for every learner. Rather, it is triggered in the mind of each learner, as part of an internal reorganization, and is based upon variables such as the following: how often a structure is needed; how badly one is misunderstood or misinterpreted when one fails to acquire the structure; how much is at stake when one misses or fails to acquire a structure; how willing one is to take the risks necessary to try new structures, etc. Since these variables are different for different learners, learners acquire structures at different times.

A second learner I'd like to examine is an Arabic speaker, prolific in his writing, who frequently uses adjective clauses but leaves the "trace" pronoun in them. For example, he'll say, I know the man who she met him. I tend to give him more credit than the Chinese speaker above, who never seemed to make an adjective clause at all, but he also is just following his native language rules about how to combine sentences. If one's native language always uses adjective clauses in this situation, then one doesn't have to master the need to use them, the point at which one decides how to put together a complex sentence, and begins to construct it. He only needs to learn how to drop out the trace "him" that sounds, to us, unnecessary and ungrammatical. Yet, if I think about it, my grading system is unfair. The Arabic speaker at least begins to make relative clauses, much more quickly, and much more easily, from a much earlier time in his acquisition. The Chinese speaker, for whom they are unnecessary, picks them up late in the game, even though constructions like the ones he makes are graded down sternly and heavily punished in formal writing. For him there is much more involved in creating them, even though he knows that this is something he should be working on.

Each speaker approaches each acquisitional reorganization as its own challenge: a reorganization that would probably be good, but also a lot of mental work, requiring a lot of attention and energy and only to be undertaken when there is enough of both. Laziness is not the only factor in resistance to change; cultural inflexibility is a strong one, and fear of loss of face is also strong. There must exist many opportunities to use the new system; the brain will not become more efficient along a single pathway, if it is not convinced that the new efficiency will help it repeatedly in the future. There must exist, at the first attempts, time to reverse one's attempts, and acceptance on the other end: acceptance of a new and unknown form; acceptance of the old form, and acceptance of the trouble caused by being forced to interpret a new version.

For a third learner, the principle of interference doesn't apply. This one is Chinese, and continually makes simple verbs with unnecessary be in them. For example, He is have a car. They are go to Chicago. This learner has also managed to work her way up to the top level classes, in part because the rest of her English is adequate; she listens well, reads well, and makes more complex sentences, yet her basic verbs are crippled by this mistaken, and unnecessary, process. Where does she get it? She has never heard nor seen verbs like this in standard English; proponents of a natural process would have to admit that, unless she was in an environment where these were actually used, actually spoken, she is not developing on a natural route. A similar process may be happening with students who take –ing forms of verbs and use them as main verbs (I sitting here) though this could be complicated by the fact that learners may actually hear these; or, as is often the case, TOEFL study before mastery of basic forms causes students to believe that this is a normal arrangement for nouns and verbs in a sentence.

If interference does not account for all learner errors, what then? The concept of interference was never to my knowledge intended to account for all errors. The method it is most closely associated with, the grammar-translation method, basically proposed taking sentences from one language grammar, and systematically translating them, so that the learner could see, repeatedly, how the native language differed from the new one; repeated and continuous comparison would eventually burn two different paths in the forest and demonstrate to the learner that the second was equally accessible, that translation is accessible and possible for all sentences. In fact most students, laboring long hours over a new language, essentially translate new sentences until they don't have to; they have this method, or process, forced upon them by default, until they are given another method. But I find myself still wondering about these learners who are left out of standard, more systematic explanations of frequent learner errors. Where do these errors come from, that do not fall under the heading of interference? What can we do for a learner, when we have no idea how they got where they are, in terms of their understanding of the language as it is? How often does this actually happen?


Leverett, T. (2008, Feb.) Communicative theory rocks the late 20th century. Google Docs. Available:

This originally appeared at and was moved here in 2010.

Work in progress; part of a larger unfinished project.