Friday, December 4, 2015

Two readables

Here are two things to look at to stir the blood.

The first is a report of a recent breakthrough for storing information on DNA. This is not a new discovery (see here), but it appears to be scaling up rather dramatically. The NYT reports the following:
The new research demonstrates that specific digital files can be retrieved from a potentially vast pool of data. The new storage technology would also be capable of keeping immense amounts of information safely for a millennium or longer, researchers said.
This is a big deal, if you keep in mind how often science follows technology. If DNA becomes the basis of long term memory storage in industry, how long before the idea hits and hits and hits again that humans do the same with their DNA. This would solve "read" part of the required read-write memory issue that the Gallistel conjecture requires. The scientists quoted in the piece acknowledge that at present the "write" part is the "bottleneck."

The scientists acknowledge that their current bottleneck is in the ability to write the information in DNA, but they say they expect that technology to begin to improve rapidly.
And when it does, we will have a realized proof of concept for the the Gallistel conjecture. And then bye bye neural nets. Can't wait.

The second piece is a blog post by my distinguished colleague Colin Phillips. He takes on the vexed issues of grad education in linguistics suggesting that it might be time for a re-think. Many of the heavyweights chime in with their views and the discussion is very informative, if also a tad defensive in parts (I won't name names, you can see for yourself). One thing that Colin does not mention, but I think is germane (and probably impolitic to mention) is that there is an understory to the debate. Like it or not, I believe that linguistics is now at an intellectual cross-roads. How so?

Part of the field wants it to stay pretty much as it has always been, and by 'always' I mean for the last several hundred years. For these the subject matter of linguistics is language and the methods basically philological. Sure GG is an important advance, but mainly because the methods developed are philology on steroids. At any rate, for this group, the mantra is "linguists study language."

The other group, of which Colin is an important leader, takes the subject matter of linguistics to be a mental faculty, either FL or G in both the narrow and wider sense (to borrow two Fitch Hauser and Chomsky terms). One studies language to study these objects that are contributing causes to linguistic "behavior." On this view, philological methods are useful but not exclusively so. This group believes that other methods both can and are contributing to understanding Gs and FL as cognitive objects.

These divergent views will take differing views on a proper linguistics education. Both will respect the standard ling methods as both agree that those methods have and continue to tell us a lot about language and G/FL. The latter however believe that other methods are no less central to understanding the cognitive (and ultimately, biological) object and so these methods should have a place in a good grad curriculum.

Now the problem, there are only so many hours in a grad education 5 years and so what to cover, given that not everything can be covered. And this is a problem where departments will differ influenced by what their view of the subject matter is. The fact, however, is that something will have to give and that allowing people to develop skills beyond the philological will require that the depths of their philological education will have to be sacrificed.  This, IMO, is a sign of progress. Grad school is there to allow neophytes to become professionals. Professionalism requires specialization. However, one cannot specialize in everything. So, choices must be made. What's the best way to make them? Well here people can legitimately disagree. However, if you are a cog-bio-linguist then you will believe that training in non philological techniques will be as legit as the other methods. This does not mean ignoring the standard methods, but it does mean making room for the others and understanding their importance for certain kinds of investigations.

Linguistics, I believe, will soon move the way of other developed sciences. Papers will be written by gangs of researchers as the expertise required will transcend the expertise of any one person. This will mean learning to work with people who know stuff you don't and that you can talk with. This does not entail becoming the other, but it does mean learning to understand the other's techniques and modes of thinking.

At any rate, take a look at the discussion. It is nothing if not amusing watching the heavyweights tussle.


  1. It's about time this discussion takes place -- but it's easy for us to talk about who have graduated at some point in the past, now have jobs, are even tenured. We can often teach what we like, we certainly can research on what we like in any way we like.

    But what about the new graduates? The future cog-neuro-bio-linguists will be even less welcomed by Linguistics/Philology departments, in many places around the world it's tough enough if you indicate any interest or expertise in GG. Ideally, there should be a new-style type of (cog-neuro-bio) Linguistics Department. But these are unlikely to be founded or (re)formed, again especially in these times of neo-liberal cut-backs and lack of vision at many if not most universities.

    Likewise, relevant graduate programs (and right now there’s only a handful that offer such expertise) will not produce too many likely hirings in CS, Biology, Neuroscience etc.

    That was just a first reaction, off to read Colin’s post. :-)

    1. In my experience, linguistics departments have not been hostile to hiring people with non-traditional skill sets, as long as those people understand and value the kinds of things already being done in the department. In fact, departments have been surprisingly receptive to this, and I haven't heard a lot of buyer's remorse. What they have been more reluctant to do is to reevaluate how well their training model fits the broadened set of faculty that they have acquired. That's the focus of my post.

  2. As a grad, I see problems too. I'm not sure that there's a one-fix approach, though I do think solutions could follow from restructuring expertise across the university. At the moment, extremely diverse interests are shoved together under the 'Linguistics' banner and it causes problems.

    My own development in cog-neuro-bio has had to be based on marginalised electives and self-study. Even where the field is utterly relied upon in the core syntax teaching and research, its theoretical foundations are entirely ignored so that everyone can just get on with typology. You'll be turned into a proficient languist with methods of analysing phrase structure, but not once will anyone rehearse the arguments in 'Aspects' for why FL must be generative and constrained.

    I will come across as an insufferable, sneering purist, but a few areas could be delegated to other faculties. Sociolinguistics, to my mind, is more Sociology > Communication than Linguistics > Society. Similarly, much of Pragmatics could be filed under Psychology > Social, and even a lot of Psycholinguistics appears to have little to say about FL.

    None of this is to say MY linguistics is true linguistics; hell, push me into CS or Biology if necessary just so we can break things up in recognition that we all need colleagues who share interests, and the possibility for that is swallowed up in proportion to how permissive your department is of divergent outlooks that are only superficially similar.

    I think this heavy restructuring would be more useful than tweaking curricula as this isn't only a staff-level problem. I not only fail to find common ground with teachers, I fail with peers, most of whom don't care about FL either. Eyes glaze over when I talk about it and it's as respectable as a linguistics grad to be ignorant of GG as it is in the public to be ignorant of basic maths. That's fine, we can't and shouldn't all be interested in the same things, but we need to herd people into more coherent groups.

    1. "I not only fail to find common ground with teachers, I fail with peers, most of whom don't care about FL either. Eyes glaze over when I talk about it and it's as respectable as a linguistics grad to be ignorant of GG as it is in the public to be ignorant of basic maths. That's fine, we can't and shouldn't all be interested in the same things, but we need to herd people into more coherent groups."

      Well put!—My experience and sentiments exactly.

    2. "extremely diverse interests are shoved together under the 'Linguistics' banner and it causes problems. (...) we need to herd people into more coherent groups."

      I don't know. Why would that be a good thing? Specialisation is important and valuable, but interdisciplinary cross-talk is crucial as well. Doesn't "herding people into groups" invite myopia and groupthink, or at least open them up to these risks more than when they'd be part of diversified faculties and departments?

    3. @mark: depends on what one means by "diverse interests," I think. I agree that we need to understand our research aims. If I am interested in the structure of FL and you are interested in typological generalizations, then maybe our interests converge and maybe they don't. This does not make mine better than yours or vice versa, but it behooves our joint interests to figure out the degree to which these pursuits should be expected to talk to one another. Intellectual coherence is a virtue even when one goes on vacation into neighboring areas.

      Now, regarding the current state of the field: I think we need to be very catholic in our choice of methods, and this suggests that certain kinds of inter-disciplinary cross talk can be invaluable. However, I also think that we spend far too much time "debating" non-issues. The case of "universals" is an excellent illustration. There is no debate here, just a pun.

      Last point: herding. If what they mean by this is having a clearer awareness of one's research objectives and questions, then I am all for it. If it is a political manifesto for dept reforms, I leave this up to individual depts and administrators. UMD has made a decision to have a cognitively oriented ling dept. Others might decide for something else. I like what we do here and can explain and defend it intellectually. I leave it to others to do the same for their depts.

  3. This comment has been removed by the author.

  4. It's interesting how much of Norbert's discussion resonates with Levinson & Evans in Lingua (the journal soon to be continued as Glossa), in their 2010 response to commentators in a special issue devoted to their 2009 BBS paper:

    "Radical changes in data, methods and theory are upon us. The future of the discipline will depend on responses to these changes: either the field turns in on itself and atrophies, or it modernizes, and tries to capitalize on the way language lies at the intersection of all the disciplines interested in human nature."

    1. I would agree with the sentiment. Where we disagree is on the substance. I've written on this and so you can see what I think in more detail if you want to. But, I agree that we should not become slaves to our methods and I agree that looking towards the cog-neuro sciences is both intellectually and politically useful (critically of course as much of the work here is not that useful given its overwhelming Empiricist orientation). In sum, L&E's meta-view seems fine. Sadly, the actual views are not so interesting.