ry-bleybaileybardovi-1983-1974-1995


 * 1) Bley-Vroman, R. (1983). The Comparative fallacy
 * 2) Bailey, N. Madden, C. Krashen S. (1974): Is there a "natural sequence" in adult language learning?
 * 3) Bardovi-Harlig, K. & Reyolds, D. (1995): The role of lexical aspect in the acquisition of tense and aspect

Consider the implications of (1) on (2) and (3). **Does one, or both, of the these papers commit the comparative fallacy?**

A summary of each paper helps answer the question.

THE COMPARATIVE FALLACY Bley-Vroman's 1983 paper is a convincing argument that the study of on the linguistic description of learner languages (e.g., Selinker's **interlanguage**) is hurt or sidetracked by comparison with the target (L2) language. This idea is called the **comparative fallacy**. Bley-Vroman focus on Tarone, Frauenfelder, and Selinker's (TFS) definition and technique for studying the systematicity of the L2 learner's interlanguage and argue that these result in incorrect assessments of the same. Bley-Vroman's claim is that the TFS approach fails because the TFS definition of systematicity does not take the structure of the interlanguage on its own terms and relies on a comparison of the target language to to study the structure of the interlanguage.

Besides their main demonstrations that the TFS technique will sometimes assess some cases of interlanguage as unsystematic when they clearly are systematic and vice-versa, Bley-Vroman detail four difficulties with the TFS framework:
 * 1) The TFS measures of systematicity do no discover subcases of **obligatory context**. That is, not all learner choices in language production are binary.
 * 2) The very TFS definition of "systematic" is susceptible to the frequency of studied word constructions in a given sample of speech it depends on a ratio of errors in obligatory contexts to the corpus. This problem is conceptual and cannot be fixed just by fine-tuning.
 * 3) See (1). The TFS systematicity measure is only well-defined for binary choices. Although binary choice can be derived via sophisticated linguistic analysis, this is not available to the learner, so for any given instance, we can't be sure that systematicity will even be defined.
 * 4) The TFS approach is at odds with sociolinguistic variation theory which directly opposes the TFS systematicity definition. In particular, variability must be measured against the internally constructed systematic rules of the learner's interlanguage, and not the the rules of the target language.

One of the interesting points in Bley-Vroman's conclusions is that if any study, especially population studies, use a target language scheme to preselect data for research (e.g., a set of errors), then they are even more apt to obfuscate understanding of the structure of learners' interlanguages.

IS THERE A "NATURAL SEQUENCE" IN ADULT LANGUAGE LEARNING? Nathalie Bailey, Carolyn Madden and Stephen Krashen (BMK), by way of their "first large-scale experimental study of the adult learner's acquisition of grammatical morphemes" (functors) seek to show that the order of learning these functors in English is invariant no matter what the L1 language of the adult. Their second hypothesis is that the order of adults' functor rankings will be more similar to those of children learning English as their second language than those of children learning their L1 language. Just like the Bley-Vroman description of the TMS technique, BMK determine the accuracy of utterances by measuring the ratio of correctly formed and used functors to the obligatory contexts for them. So BMK's analysis is still comparing the learners' language to the target language in a manner similar to the TFS approach and is thus exposed to the same difficulties and incorrect and inaccurate assessments of the comparative fallacy highlighted by Bley-Vroman. **So yes, this paper commits the comparative fallacy.**

However, even with this said, there are some good observations in the paper. For example, adults clearly benefit from formal instruction and do not benefit much from simply being exposed to informal English speaking environments. Also, even though adults are taught English functors in orders different than the "natural sequence," they still profit well from the instruction. I wonder as I imagine the paper's authors must have wondered, whether teaching the functors in the natural order is significantly more efficient and worth the corresponding syllabus adjustments. I know that in the typing world, the notion that the Devorak is so much better than the QWERTY keyboard turns out to be a myth based on intentionally biased research: [|Typing Errors on reason.com (1996)].

THE ROLE OF LEXICAL ASPECT IN THE ACQUISITION OF TENSE AND ASPECT The previous paper and this paper by Bardovi-Harlig and Reynolds were good to read one after the other because the contrast shows how much better the SLA theory works in the latter. The comparative fallacy is not committed in this paper because the authors are not comparing the learner language with the target language in the old contrastive analsysis way. Instead, the authors have analyzed the lexical aspect, as in the expression of action and time of L2 verbs in the simple past tense of L2 language learners. It turns out, based on their research on larger numbers of L2 learners at various stages of L2 learning, that L2 learners do have a preferred order based on lexical aspect in when they are learning the simple past tense. Their results show that learners use event verbs (achievements and accomplishments) as best case examples for past tense and use past tense less with activity and state verbs. What is very interesting, is that this is true both for classroom taught learners as well as "untutored" learners. The authors speculate that the influence of lexical aspectual class may be an **acquisitional universal**.

The other item that the authors that has an impact on their L2 teaching advice is that both untutored and classroom-taught L2 learners seem to follow the **one-to-one principle** between the form and one meaning. That is, they under-generalize in their learner grammar. The big deal in this paper (for me) is that besides not committing the comparative fallacy because they are relying on studying the learner grammars independent of their L1 languages, they make some sharp observations about focused noticing and positive evidence. They suggest that by getting learners to notice differences between their (interlanguage) grammars and the target language grammar, learners will be able to efficiently continue learning even outside the classroom. They will continue noticing particular differences and be able to make adjustments because they will be getting positive reinforcement via authentic and naturally occurring language input.