Krashen - The basis for our framework

Stephen Krashen is an American linguist at the University of Southern California. His theory developed in his book "Principles and practice in second language acquisition" (Krashen, 1982) is one of the major works in the literature on language learning. It has served as a basis for research papers from many other authors as well as Krashen's own further work and is still being discussed today. It provides us with the basic from which we will derive our initial framework on how to learn a language. Krashen starts by stating that there is a disconnect from the science of how to learn a language and the teaching practices employed by teachers. This, however, Krashen says, is not because of neglect from the teachers, but rather because the science community has failed to communicate its practical advice and also changed, which means, that the advice teachers got in the past was not as mature as it is now. The past theories focused on the wrong things, Krashen explains, such as the 'applied transformational grammar' theory, which focused on the result of the learned language instead of the process of acquiring it. Teachers thus have lost some degree of trust in the science behind language learning and instead rely on their intuition and anecdotal evidence of which teaching methods work and which do not (Krashen, 1982, p. 10-11). Krashen proposes a new theory of second language acquisition. What makes it different is, that it tries to explain the process and not the result (in form of developed language competency) of language acquisition.

The acquisition-learning distinction

Krashen makes a claim he is well known for: There are two distinct and independent ways of developing second language competence: acquisition and learning. Acquisition is subconscious, which means the learner is unaware of being in the process of acquisition or of the result of having made the language acquisition. After acquisition, the learner feels that some language piece is correct or wrong, but he doesn't know why this is so. The only thing the learner is aware of, is when he uses his L2 competence for communication. Learning and the resulting knowledge, in contrast to acquisition, are conscious. The learner has explicit knowledge of the L2 in form of grammar or rules. Error correction has no effect on the subconscious acquisition and only has an effect on the explicit learned knowledge. Its overall impact is questionable as we will soon discuss (Krashen, 1982, p. 13 - 14).

The natural order hypothesis

When acquiring a language, the sequence in which the grammatical structures are acquired by a user seem to be in a predictable order. Some grammar rules are learned early in language development and others are only learned near the advanced stage (Krashen, 1982; Truscott, 1996). This order is not identical for every user and for the first and the second language. However, there are significant similarities. Krashen adds that further studies have confirmed this order in other languages. Finally, Krashen concludes that when acquisition is the goal, grammatical sequencing when teaching should be avoided (Krashen, 1982, p. 15 - 17).

The monitor hypothesis

The acquired and the learned language competence are used in two different ways, Krashen claims. The acquisition system provides the user with fluency. When the user wants to produce communication, that spontaneous result is provided by the acquisition system. The learned system then can monitor the product and detect any potential errors, which then can be edited to correct. This, Krashen states, can happen before or after the actual output, which would be speaking or writing. The very limited usability, as described below, of the monitor implies that it has a very minor role in L2 competency (Krashen, 1982, p. 18 - 19). Krashen presents three conditions which are necessary but not sufficient for monitor use, meaning that even if all three are meet, the user can still decide not to use the monitor:

Time: the user needs sufficient time to use the monitor and edit the product. In most language communication situations however, time is not available and the overuse of the monitor in such situations can be bad for the fluency of the user.

Focus on form: For applying conscious grammar rules, the user has to consciously think about producing correct language. However, in real communicative situations it is often the case that the user is focused on the message or content rather than the correct form.

Knowing the rule: Even linguists don't know all the grammatical rules since language can be infinitely complex. This means, that the user is only exposed to a small subset of rules and even the best students don't learn all the rules that were made available to them.

Even with these three rules it is difficult to get the user to use the monitor. Asking him explicitly will still not make the user use the monitor. Krashen explains that this is shown by the users error pattern's, which will still reflect the natural order (the sequence of grammar acquisition) and he will still make errors that he consciously knows to be errors. Only specific grammar tests that only test's for the learned grammar will reveal the learned system in the user (Krashen, 1982, p. 18-19).

Moreover, Krashen suggests that, the monitor is best used for simple grammar rules that are on a low level cognitively. If the rule is simple enough, the language product might even be edited in real time.

Krashen then presents three different types of monitor users that differ in their monitor usage. There are over, under and optimal monitor users which represent their intensity: Monitor over user: These users try to always use their monitor, even to the point of frequently interrupting themselves in order to self-correct. This use type will likely impede their fluency even if they could be fluent. Monitor under users: This type doesn't learn and if he does learn he doesn't use his learned system. He is uninfluenced by error correction and uses his acquisition system to self-correct by 'feel'. Optimal monitor user: The user uses his monitor when it is appropriate and when it doesn't interfere with his communication. This is usually the case when he is writing, since there is sufficient time. By doing this, the user has found the optimal monitor use as he is complementing his subconscious acquisition system with his conscious learned system to fill the skill gap of his incomplete language competency. This means, that the user still makes many small mistakes but can give the illusion of being a native speaker in his writing (Krashen, 1982, p. 20 - 21).

The input hypothesis

This hypothesis lies at the heart of Krashen's theory. In the traditional pedagogical method of language teaching a deductive approach is taken. First the grammar structure and rules of a language are laid out, then for every grammar structure there are exercises to drill down that skill, thus developing fluency by practice.

Krashen's input hypothesis runs counter to that. His approach bears similarities to an inductive method: There should be comprehensible and extensive input first in the form of extracting meaning out of communication. Doing this, over time, reveals the overlying grammatical structure which is then acquired by the user.

To make how acquisition work more explicit, Krashen investigates the question: How can a user advance from his current level i to his next level i + 1? Krashen states, that he does so when he is at stage i but understands i + 1. Krashen points out that this doesn't seem possible, since if the user is at level i, how can he understand level i + 1? The reason this is possible, Krashen explains, is that 'understanding' refers to the message and not to the syntactic form of the message. This means the user can utilize extra-linguistic information to understand the meaning. This extra-linguistic information can be anything. When reading an article and not understanding some words, this might come from the context of where the article is from or what it is about. It can be the general context of the situation such as the weather. When paying for food at the supermarket, it is possible to communicate without exchanging a word. Finally, its is possible to communicate on some level just by body language alone.

Krashen claims that, for the input to be acquired, it has to contain i + 1. However, Krashen adds, that it is unnecessary and can even be harmful to deliberately provide i + 1. Instead, as long as the user understands the input, or the communication is successful, i + 1 is automatically provided, Krashen states.

Krashen warns that output (speaking or writing) should not be focused on or even forced externally. Instead, he explains, when time has come and the user feels ready, fluent output will slowly emerge. Subsequently, fluency cannot be taught directly. The emergence is different for every user. This output will not be grammatically correct and will contain many mistakes, over time and after more input, however, there will be less and less mistakes.

Krashen condenses the input hypothesis with these four statements:

"1. the input hypothesis relates to acquisition not learning [...]. 2. We acquire by understanding language that contains structure a bit beyond our current level of competence (i + 1). This is done with the help of context or extra-linguistic information [...]. 3. When communication is successful, when the input is understood and there is enough of it, i + 1 will be provided automatically [...]. 4. Production ability emerges. It is not taught directly." (Krashen, 1982, p. 22f)

In addition, Krashen gives real world examples of interactions characterized in his input hypothesis. In the real world, when people, who have not developed their language competency to native level, communicate with each other, they modify their language, which Krashen calls 'modified talk'. When interacting with L2 beginners, others will modify their talk to provide the beginner with modified input that roughly mimics the i + 1 of the L2 beginner. This happens, as Krashen notes, almost automatically and subconsciously with foreigner talk, language teacher-to-student talk and intra-language talk meaning both speakers are L2 beginners.

The input hypothesis also indicates, Krashen says, of why a grammar focused syllabus might be of limited use: When teaching in a 'grammar structure of the day' style, all students need to be on the same stage in their acquisition system and learned level. This, however, is contradicted by the i + 1 of the students being potentially different. With a communication focused teaching style, Krashen suggests, that in most cases, this provides every user with their personal i + 1. Moreover, Krashen notes that the grammar focused teaching is not flexible as every rule is only presented once, and at most to be reviewed later in a rather artificial way. If the users are not fully attentive to what the teacher is presenting, they will miss out and fall behind. This limitation, Krashen claims, is again elevated by a communicative teaching style because the provided input contains a natural review of the grammar structures that appear in it. Finally, Krashen points out, grammar focused teaching relies on the assumption that the natural order (sequence of grammar acquisition order) is known and this assumption is not needed for communication input that contains i + 1. Krashen warns that teaching focused on grammar is hard to combine with teaching focused on communicative input because the moment input is used as a vehicle for grammar, its message looses its authenticity and thus becomes meaningless to the user. Subsequently, grammar takes time and focus away from communication, so it should be avoided (Krashen, 1982, p. 26).

Krashen concludes the input hypothesis with the prediction that the classroom will be a good environment for L2 input-focused acquisition because, compared to the outside world, it can provide much comprehensible and targeted input for users. The classroom thus is best suited for beginners up to the intermediate level (Krashen, 1982, p. 21 - 26).

Note that Krashen states that he views both the inductive and deductive approach as 'learning' and warns about mistaking his 'acquisition' for an inductive learning approach. His reasoning is, that inducing rules from input is still conscious or of the same learning nature as opposed to acquisition (Krashen, 1982, p. 99f). While one can agree with this to some degree, inducing such rules is not inherently conscious as they might be emulated and of the same subconscious nature that acquisition is.

The affective filter

The affective filter explains the impact certain affective factors have on L2 acquisition (Krashen, 1982, p. 30). Krashen points to three types of affective factors which help in the success of L2 acquistion: Motivation, mostly integrative (similar to intrinsic) but also instrumental (similar to extrinsic), self-confidence and both low personal and classroom anxiety. This is in line with what we already have established in the previous part about motivation and emphasizes the role that motivation, more generally, has. Krashen claims that these three types only influence the acquisition and not the learned system. He states that this can be seen in the fact that they predict improved test scores in tests for acquisition, but have no effect on grammar tests.

Users vary in the strength of their affective filter. This partly explains why some users stop learning and fall short of native level competence, even if they are exposed to a high amount of comprehensible input. Krashen states that a high affective filter will cause a user to seek less input and when being exposed and comprehending input, the filter will block this input from the acquired system in the users brain, thus disabling acquisition. Krashen contrasts this with a user that has the optimal attitude and mood, who will be predicted to have a lower affective filter. Such a user will actively seek out input and when comprehended it will deeply reach his acquired system. As a consequence, a classroom that doesn't put students in the defense and creates a low (or no) anxiety environment is predicted to lower the affective filter and be more effective, Krashen claims (Krashen, 1982, p. 31).

The role of error correction

John Truscott (Truscott, 1996) argues that grammar correction in writing classes should be abandoned. He suggests that it has no effect except that it might be harmful. He adds that teachers are hesitant to accept this, taking its value for granted: "teachers (and nonteachers) show great reluctance to accept [...] that grammar correction does not work" (Truscott, 1996, p. 341). He suggests that an underlying reason might be the misconception teachers may have that their goal is a 'transfer of knowledge' from teacher to user, underestimating the complex nature of building language competency. Here Truscott proceeds to refer to the natural order, which Krashen was also referring to: The acquisition of grammar is gradual, not instantaneous and users are unlikely to be at a state within their natural order to be ready and open for a respective grammar correction, resulting in correction being a distraction to the user. Truscott further states that if grammar correction has an effect, it is most probably just 'pseudolearning' which results in 'pseudoknowledge' (Truscott, 1996, p. 345). This is when the teaching system fails to access the desired competence area. The result of this is superficial knowledge, which has almost no use to actual language competency. Moreover, when users are able to detect and correct errors, they are still unable to name a rule or explain the error. In addition, users tend to follow their intuition instead of their explicit language knowledge, making mistakes that they would know not to make and ignoring previous teacher corrections. When users react to correction, they commonly do so by simply avoiding the situation in which the error could occur and revert to write overly simple (Truscott, 1996, p. 349).

Here Truscott closely matches Krashens learning-acquisition distinction where learning would be pseudolearning and acquisition would be learning that has utility for language usage.

Truscott also notes that teachers tend to be inconsistent in their correction, that they might not know the correct from or how to explain it, that they very well might be too busy, that the user might not understand the correction and finally, that correction takes time away from other more valuable activities (Truscott, 1996, p. 350).

Truscott also warns that grammar correction can be harmful insofar that users can find extensive grammar correction "extremely discouraging" as they don't like to be shown wrong multiple times (Truscott, 1996, p. 354). However, he acknowledges that users might still want to be corrected as they might think it will benefit them because they don't know how useless it is. The proponents of grammar corrections argue that it is required to avoid 'fossilization': This refers to users not learning the correct or learning the wrong form and then becoming stuck with it. Truscott counters this by saying that there is no evidence to support this claim (Truscott, 1996, p. 357). Finally, he argues that users might have wrong intuitions about grammar corrections and thus insist on it. Instead, teachers should re-educate them and still refrain from it even if it goes against the users wishes. This seems to be in conflict with maintaining the users autonomy. Instead, it might be beneficial to inform the user (as Truscott suggests), but to still offer correction in the from of feedback as an option (if offering it doesn't take away resources somewhere else). After some time the user might confirm that he doesn't need it by himself.

While Truscott's claims are in relation to grammar correction in writing classes specifically, the underlying insights might be generalizable to general practice besides writing.

Conclusion about Krashen

Krashen summarizes his theory with the following statements:

"1. Acquisition is more important than learning. 2. In order to acquire, two conditions are necessary. The first is comprehensible [...] input containing i + 1 [...] and second, a low [..] affective filter [...]. We may see positive correlations between other variables and measures of achievement in second language, but in all cases in which language acquisition is attained analysis will reveal that the relationship can better be explained in terms of comprehensible input plus filter level" (Krashen, 1982, p. 31f)

While the theoretical background on motivation provides us with a good starting point on why a user might want to learn a language, Krashen's theory gives us insight that we can use as a guide. His main point is to move away from 'tedious' learning and focus more on acquisition. Acquisition could be described with 'learning by doing' because the user acquires with comprehensible input that focuses on communication and not on syntax. In other words, the user just has to use the language as it is intended to be, as a communication device. This is in agreement with our conclusion from motivational theory, in that people are motivated to learn and use languages primarily for the purpose of communication and understanding of other cultures.

Nonetheless, one major question still remains. It is reasonable that the user acquires new words by input, namely by reading text. This, however, requires already preexisting knowledge. If i is the number of acquired words and has the value 500, i + 1 works well. But what if i is zero? How does a user build up his initial i? In other words, it is unclear how the user should learn his first vocabulary words. Is learning this vocab indeed as useless as Krashen makes it out to be? And if initially learning vocab to build up a word bank, by repetition and drill, should this initial learning part be abandoned? Has learning no place in acquisition?

We will attempt to answer these questions soon, but first, we have to look into a more fundamental theory that will help us to approach these questions.