Must a person speak a language to understand it?

This is an important question to ask for instructors of ancient languages who want to move toward communicative approaches to language instruction, and the answer to this question is: no.  This doesn’t necessarily rule out that we should use communication in ancient language instruction (though for this and a number of other reasons I wouldn’t), but it should at least mitigate the degree to which instructors insist on it for everyone.

One problem with the view that a person must speak a language in order to properly understand it, which we sadly learn about mostly from the misfortunes of people with language disorders, is that language production and understanding dissociate.  In other words, different language problems can affect either production or comprehension.  I was reminded of this while reading Gregory Hickok’s The Myth of Mirror Neurons shortly after writing my previous post which pertained to some extent with language instruction and understanding.  Hickok describes at length the observations of Eric Lenneberg on one of his patients:

Lenneberg noted that the child’s ability to speak was effectively nil. The child cried and laughed normally but made only cough-like grunts that accompanied his attempts to communicate via pantomime. When he played alone, he often made sounds that resembled Swiss yodeling, although he had never been exposed to the art form. Lenneberg further noted that as the child got older, and after years of speech therapy, he could repeat a few words after his therapist or mother, but even these attempts were “barely intelligible” and were never produced on his own.

Lenneberg was fascinated by the boy’s ability to comprehend language, which the scientist characterized as “normal and adequate understanding of spoken language.” This observation was confirmed over more than 20 visits and by several doctors and speech therapists, both informally and formally. The boy could follow complex commands such as “Take the block and put it on the bottle” and he could appropriately respond, albeit nonverbally, to questions about a short story that was read to him. He was not merely cuing off the body language of the researchers: he accurately followed commands even when he couldn’t see the investigator. The cause of the speech production deficit could not be determined. Lenneberg called it “congenital dysarthia” – difficulty controlling the speech articulators – and concluded that the ability to master receptive language, to understand, did not depend on the ability to produce or imitate speech.

Dissociations between production and comprehension are also present in different kinds of aphasia (for good introductions to aphasia see Basso and Goodglass; Basso is probably the better of the volumes, but Goodglass is a little easier to read). Aphasias also differ in how they affect the ability to read.  Yet Lenneberg’s patient and others like him demonstrate that a person can understand a language though they may never have been able to produce it. Of course, this should be taken with a bit of a grain of salt since, as I pointed out in my last post, a second language will never be the same as a first language.  But, even working on that analogy, it does not seem to be the case that the ability to speak a language is a necessity for understanding it.


  • I am sympathetic to the arguments from the communicative side, as their goal is to move beyond a decode/re-encode model of reading the second language. Having said that, the question remains whether this is necessary or simply sufficient, and the jury is still out on this. We have a huge cultural chasm to overcome in terms of cognitive lexicography, one which a communicative approach will not overcome any more than decode/re-encode. Speaking might help one understand the pragmatics of intonation and information structure, but huge gaps will still remain. There is no single solution to the problems. I hope that we can move beyond the current either/or debate to identifying the specific problems, a to-do list if you will. Thanks for the post.

    • Thanks for the reply Steve. I think you are right that one of the biggest matters is to get past either-or. I would comment, though, that I don’t share as much the concerns about decoding/re-encoding, nor do I think many instructors of modern languages, since language learning is not an all or nothing process. For example, many applied linguists view moving from passive recall (the ability to recognize a second language word and give a first language “equivalent”) to active recall (the ability to recognize a second language word and give a second language “equivalent”) as sliding scale or process. This is the motivation behind things like the Before You Know It flashcards created by the psycholinguistics who are partly responsible for designing courses for Transparent Languages. There was actually a study done back in 1994 or 1996 by Laufer and Goldstein, I believe, though I’d have to look it up, that demonstrated that passive recall was the best predictor of overall language learning success. So, I think it still remains to be shown that decode/re-encode is really as much of a problem as some would suggest.

      • I should have been more specific about what I meant by decode. In working with NT students, many are translating phrase by phrase without a framework for moving above that level, leading to fractured understanding. The communicative approach can help with this, but I’m not saying that is the only way. One of the goals of the LDGNT/LDHB project is to provide an aid in this regard, but there is still much more that has been left out. The decode/re-encode teaching I’ve experience has focused on translation rather than understanding. I know the latter is fuzzy to assess, but it still needs to be the goal. I think this is what has led to translation as the pinnacle of success instead of comprehension and understanding. The profs I studied with presupposed that the ability to translate meant the student understood, but this was not the case for many. I am on the decode side as well, but not as it stands. We need more tools and resources to move beyond where we are, but we also need to spec out the problems that need addressing.

        • Ah. Yes, with that I would be in total agreement. In the program at Stellenbosch, meaning is the centerpiece of instruction as opposed to translation, though that’s a component. Comprehension of texts is also stressed as opposed to only phrase by phrase understanding. So, I do think we share goals with those moving toward communicative approaches, only we think it remains to be shown whether communication is a necessary component, or at least whether it must be as central as some claim.