# Does the Sapir-Whorf hypothesis apply to artificial (specifically programming) languages?

As a computer programmer, I have an intuition that the idea "language influences thought" is very relevant to programmers and programming languages.

• Is there any research that examines whether artificial languages influence thought?
• How can such a theory be expressed as a testable hypothesis?
-
NLP as in Natural Language Processing is also not an appropriate tag for this question. If by 'artificial languages' the OP means 'programming languages' and how they effect how we design/implement/reason-about algorithms, then he should write programming languages, and the io-psych tag would be relevant. But usually 'artificial languages' means languages that are created artificially for normal conversation (say Klingon, or Esperanto). Which do you mean @Roly? –  Artem Kaznatcheev May 25 '12 at 16:28
@ArtemKaznatcheev It's possible that I missed the mark with that one. I know for a fact that it wasn't neuro-ling-prog, so perhaps linguistics alone (or with a new tag) is most appropriate. Great question. I don't think IO plays into this at all. –  Chuck Sherrington May 25 '12 at 16:37
@ChuckSherrington if this was a question about programming languages wouldn't that relate to io-psych? Since organizations would be interested in optimizing what tools they give their programmers? We can discuss this in detail in chat. As is, I think the question is fine with the linguistics and maybe language tag. Here are 1, 2, 3 related questions; we could steal some tags from them. I recommend philosophy-of-mind. –  Artem Kaznatcheev May 25 '12 at 18:04
@ArtemKaznatcheev Sorry for the delay in my response, I did have specifically programming languages in mind when I wrote the question. Thank you for the interesting discussion. I wasn't really sure what the most appropriate tags were. –  Roly May 27 '12 at 22:00

Firstly realize that Sapir-Whorf was proven wrong in the strong sense, but is accepted in the weak sense. I don't think there's really any doubt now that learning new symbols and languages, or just enhancing your vocabulary in one language, influences thought. Many mathematicians are especially proud of how accurately proofs communicate, and having gained this accuracy, the mind comes to expect it more and loathe ambiguities. You see this happen with a lot of people in quantitative disciplines.

There doesn't seem to be a lot of research, but you're not the first to think of it. Many programmers before have too (see wiki's references):

http://en.wikipedia.org/wiki/Linguistic_relativity#Linguistic_relativity_and_artificial_languages

Of course, none of them were neuroscientists or psychologists. To test for causality would be difficult to control for (you'd have to provide a mechanism and we're only just about getting there at the neural systems level). On the other hand, I'm sure you would find a correlation between programmers and certain kinds of behavioral and cognitive traits very easily.

In the end though, I don't really think there's any doubt among neuroscientists or psychologists; it's just a matter of finding the neural mechanism. Language has been a pain in that regard.

-
+ 1 Good answer, but why is causality difficult to test for? You can study child behavior before they learn to program and then after, and have a control group that isn't taught to program. Sure it doesn't uncover a neural mechanism but that isn't necessary to establish causality in psychology. –  Artem Kaznatcheev Jun 17 '12 at 14:48
I would only call that a correlation without a mechanism. The mechanisms doesn't necessarily have to be neural, but the mechanism should have external validity. –  Keegan Keplinger Jun 18 '12 at 19:21