In considering how to respond, Fish points out several old argument that won't work.
Well, it won’t do to invoke the pieties [that] ... the humanities enhance our culture; the humanities make our society better — because those pieties have a 19th century air about them and are not even believed in by some who rehearse them.
And it won’t do to argue that the humanities contribute to economic health of the state — by producing more well-rounded workers or attracting corporations or delivering some other attenuated benefit — because nobody really buys that argument, not even the university administrators who make it.
I think Fish is correct in saying that these arguments won't work, and he resigns himself to the possibility of politics, or of a limited few powerful people pushing some important buttons since they have a personal value of French, or theater. But I think the rest of us should not breathe a sigh of relief, but attack the assumption that these programs are less necessary than ours. This logic will quickly lead to our own doorstep, because most of us do not have a firmer foothold than theater, or french, or russian when it comes to direct economic utility, or contribution to society or culture. When you consider where these arguments take us, you quickly come to the conclusion that people should be in professional training programs as soon as humanly possible. Why waste time studying y if you know that you are going to do x?
But, as the title to this post declares, I think there is a strong case for studying many things, including theater, French, and classics. I believe this case is first made by several studies which I will outline below. But further, these studies (and many others) should urge us to be humble in the face of our increasing drive towards narrow training at the cost of education, and towards applied pursuits in the search of a specific goal at the cost of basic intellectual inquiry in pursuit of the pleasure of knowing. This trend of more training and less education is pervasive in our current educational system, and can be seen in K-12 reforms like NCLB and RTT (which evade politically controversial curriculum changes, but end up coercing teachers to become reading "trainers" rather than seeking to instill a love of reading and knowledge) to other accountability measures, coming soon to a college near you (paywalled Chronicle of Higher Ed piece, but you get the idea).
So, what is the scientific case for French or Italian or Russian? First, there are diverse cognitive benefits for bilingualism. Ellen Bialystock's work has documented that bilinguals have beginning troubles with the competing languages, but that this leads to very long term and general benefits for what is called executive functioning (brief Washington Post summary) which generally concerns distributing one's mental resources. Bilinguals are therefore more able to ignore distracting information, even in some basic, non-language tasks. Her recent research suggests that bilingualism can delay the onset of dementia, by an average of 5 years (any economist want to calculate the cost to society on 5 years of dementia?). Further, it seems that learning two (or more) languages can enhance our concept formation and cognitive flexibility (when we understand how words can have subtly different meanings in different languages, it illuminates the flexible nature of language itself) (short blurb here).
Second, there are social, cultural and ethical advantages to studying a foreign language (especially study abroad). Yes, the students love it (but they also seem to drink and party more, so that is no surprise. But study abroad programs also enhance creativity (this link to the original article in a psych journal probably won't work). Study abroad also enhances cross-cultural tolerance and a global awareness. This tends to be a goal of a college education, but it can't be done by tolerance seminars, or even hundreds of generic exhortations. You can't just learn to be generically tolerant, you have to learn a particular culture.
Finally, I think a take-home message we should all get from the science of why there is value in the humanities (and the liberal arts in general) is that we should be humble in our drive to tie education to specific and direct goals. This approach is short-sighted, not just because bilingualism improves creativity and prevents cognitive aging, but because most of the effects of any sort of education are very very hard to measure. We psychologists can assail education research for not providing clear answers on anything, but at some point we have to conclude that the kind of clear answers we want just don't exist. Assessing the independent value of a good kindergarten experience (for example) is incredibly difficult, if not impossible. But in our striving for accountability (such a reasonable sounding goal), we are increasingly narrowing our educational goals to those that are easier to measure. This first drives out the humanities (theater! how do you measure outcomes of that?) but eventually it will drive the mind out of the academy and make trainers of us all. And ironically, I think we'll find that the job training and all those 21st century skills didn't turn out to be "trainable" skills at all, but depended on the broad body of knowledge that we have been working on for over 200 years.