Corder’s model was the first model to try to illustrate the development of interlanguages and the stages they go through. It does not imply however that all language learners go through the exact same stages, nor does it imply that any two learners in the same stage will even have very similar characteristics. It is simply a model that tries to give a very general picture of the development process. Because its aim is to give a general picture, it makes certain overgeneralizations about the process, and may be misleading if taken too literally.

Corder’s model attempts to generally describe the development of interlanguages, but does not take into account the separate development of individual skills or even specific rules, nor does it take into account the immense differences that may exist between any two language learners.

The following summarizes some of the characteristics of the development process that the Corder model does not account for:

• Different skills aren’t necessarily developed simultaneously. For example, a learner may develop her interlanguage ability to speak, and this ability may grow and even stabilize without her ever learning how to read. This may happen in the case of uneducated adult immigrants, or in children who are too young to learn how to read. For this reason, we cannot assume that a learner in a higher stage is necessarily stronger than a learner in a lower stage in all skills. Although the language learning process is very systematic, it also varies a lot from one learner to the next.

• It also needs to be pointed out that learners are not learning all the forms of the target language simultaneously. Most of the time, learners will make progress in one form while others wait. In some cases, learners might completely master one form before ever beginning to learn another. Therefore, it is not safe to assume that a learner in the systematic stage knows and uses every rule systematically. Nor is it safe to assume that a learner in the random stage has no form that she uses systematically or automatically. It is necessary to be aware that some forms get acquired systematically and automatically long before other forms ever get learned in any way at all. Thus the Corder model does not account for the developmental stages learners pass through while acquiring individual forms, or the natural order of acquisition that constrains their acquisition of certain forms. In reality, it may in fact be impossible to acquire the systematic use of certain forms until certain other forms have been acquired systematically and automatically.

• Taking into account the last two points, it is imperative that one realize that it not safe to equate the stages with linguistic competence, or "accuracy". The model only proposes to show interlanguage development as a continuum from random to systematic but does not state that systematic use of a language is equal to accurate use. Some learners may use their second language very fluently and systematically but still retain many inaccurate forms or even be lacking many forms. This can be seen in the cases of learners who are stabilized while still incomplete, or "fossilized". Taking this into account, it is not safe to assume that a learner in the random stage is a true beginner, or that a stabilized learner has advanced communicative competence.

• The Corder model also does not take into account the difference between explicit and implicit knowledge. It is possible for example, for some learners to have studied the target language for a long time and actually know a lot about the language without ever developing any ability to use it systematically, automatically, or fluently. In that sense, a learner’s knowledge about the target language won’t get them to a higher stage unless that knowledge somehow becomes internalized through frequent use. Likewise, it is also possible that a learner is in the systematic or stabilized stage, but completely lacks any sort of explicit knowledge about the language (as in the case of a child in an ESL situation or an adult who learns the language in a "natuaral" way). Once again, it is not safe to assume that a learner in the random stage is a true beginner.

• It can also be misleading to overgeneralize Corder’s notion that learners in the systematic stage can correct their own errors. For example, most learners in the stabilized stage are fossilized, and therefore still retain some incorrect forms. When asked to correct their own errors however, they cannot. They may not even know what is wrong. In one case for example, a stabilized learner might produce a sentence such as "I saw the man he stole the bag" because she has not learned how to produce a proper relative clause (e.g. "I saw the man WHO stole the bag"). When prompted or cued by a teacher to correct herself, she may not be able to do so without help. Corder’s idea is not unfounded though. Corder chose to equate self-correction with the emergence of a rule (i.e. the emergent stage) because of research data. In research, it was found that children begin to self-correct a specific rule shortly before acquiring it. This is true for first language learning. Notice that this ability is for specific rules however – the rules that they are about to acquire. It cannot be generalized to the whole system though. It would be an overgeneralization to say that a language learner begins to acquire the ability to self-correct every form all at one stage, just as it would be an overgeneralization to say that learners acquire every rule all in one stage. Just as different forms are acquired at separate times, so too may this ability to self-correct those various forms. It is also not logical to assume that the ability to self-correct a form necessarily implies that the learner will soon acquire that form. There are certain grammar forms that are quite "simple", and therefore, easy to self-correct. Yet, those grammar forms may take a very long time to acquire systematic and automatic use for (e.g. plurals, simple past-tense regular verbs, subject-verb agreement). In other words, adult learners may be able to self-correct certain forms while still in a stage in which their use of those forms is quite random. Adults ability to self-correct may be attributed to explicit knowledge of those forms and not to emergence of those forms.

For all these reasons, it is better not to take the Corder model literally. It would be better to remember that it is just a model and therefore has flaws typical of any model. It may be impossible to represent something as complex as language acquisition in one simple model. It is helpful nonetheless, and, relatively accurate when we focus on the main ideas.

The following are some of the main ideas that the Corder model expresses well:

• Interlanguages are in works progress. They are moving forward, gradually becoming closer approximations of the target language. They are moving away from the first language, eventually becoming totally separate systems not dependent on translation or first language transfer – thinking directly in the target language.

• Because interlanguages are integrated systems of knowledge, their development can be described as the development of systematicy. At first, the learner’s knowledge or ability is not internalized, so it is not systematic or automatic. Eventually, as the system emerges, it becomes easier for the learner to use the language. This gives the learner fluency but does not imply accuracy. Interlanguages are systematic, but their system is often idiosyncratic. Accuracy may increase over time, but the model is not about accuracy.

• As the interlanguage develops the learner relies less on her first language or her teacher for new knowledge, and begins to discover the rules for herself through the act of using the language.

• Ultimately the forward motion is made possible by the internalization of the learner’s knowledge. Internalization of the knowledge takes place as the learner uses the target language. It comes from experience.