How long will there be computer science departments? | Hacker News Reader

How long will there be computer science departments?

johndcook.com

60 pointsfogus42 comments

How long will there be English departments? Every professional needs to be able to communicate ideas in writing. One would assume that in time English would be absorbed into other departments that applied it.

I'm equally confused by his point about non-statistics departments doing more statistics than their "proper" counterparts. While that may be true, the bulk of cutting edge research in a given field always comes from specialists. I've never heard of substantial statistical advances arising in psychology departments (even though they're arguably one of the biggest "consumers" of statistics out there).

It's a bit like trickle-down economics.

I think you have to look 'up' the stream. Physics has given statistics some of the biggest ideas in the recent past. I think biology is relatively mathless unless it's *omics crap, and then again you have non-statisticians contributing to stats. Also plenty of CS stuff for stats.

There are definitely other mathy areas: crystallography, biophysics, computational neuroscience, theoretical ecology, and applications of simulation methods like CFD and FEA to physiology, . . .

But yes, most biological research doesn't require much mathematics.

can you give some examples of statistical advances from physics?

Econometricians (based in economics departments) have contributed to the cutting edge of statistical knowledge.

Financial econometrics, in particular, is among the most adavanced actual uses of statistics. This is also not all occuring in academe.

meta analysis?

http://en.wikipedia.org/wiki/Meta-analysis

Invented to look at telepathy, and Rhine was definitely a psychologist (though one the field unfairly tries to forget). That being said, I thought Glass was a psychologist but he was actually a statistician, so I suppose the modern form of meta-analysis was, in fact, invented by a statistician. Do psychometricians count as statisticians? If not, then quite a few more things were invented outside of statistics.

Psychology: not just for innumerate touchy feely hippies (just mostly).

Except that English as a subject is much less about writing English or the English language as it is about studying English literature and its reflection of (and impact on) culture and history.

Depends on which English degree you study. I know those who've studied the literature and cultural impact side of things and those who've studied the language with a linguistics spin on it.

http://www.paulgraham.com/essay.html

Has an interesting history on the subject

Some writing courses have been absorbed by other departments. Engineering departments, for example, may teach their own technical writing courses rather than going through the English department. When I was in college, I had a writing course offered through the physics department.

> How long will there be English departments?

For however long there will be people willing to pay tuition to enroll in one. Really, that's the answer to all questions in that form. No matter how worthy of independence a study might be, it exists or fails to exist primarily on the existence of a body of students willing to pay money to sit in those seats.*

* speaking purely in terms of undergraduate programs here

Damn it I came to make exactly the same comment.

Writing English stories is just aiming at a different compiler in someone else's head.

"Writing English stories" has pretty much nothing to do with what people do in English departments.

I doubt they sit around writing all day - but I am darn sure they do compiler anaylsis of the human brain.

So they finally got rid of those pesky creative writing classes, huzzah!

From the comments on the linked article's site:

>I currently work in a bioinformatics group, and have been in computer applications for science for 20 years. I do not think Bioinformatics will be going back to biology, I see it as going back to computer science and biology. As the tools get better, there will be less and less need for people that understand both – there will still biology computer applications, but it will be commercialized and individual labs will not have their own developers. You can already see the trend starting. Statisticians do not write their own stats programs, they buy SAS or download R, in 10 years biologists will be doing the same.

I tend to agree with this.

More than computer science departments going away, I see other departments following the lead of Schools of Business Administration. Essentially every SBA now offers a degree that is essentially a 'Management Information Systems' degree, which instructs on how to take the specialized applications that exist out there and leverage them in your firm to accomplish any number of organizational goals.

I could see a trend where more and more departments develop their analog to this, suited to their specific discipline.

I think that, though computer science has permeated a number of different fields, its core study remains more relevant than ever. There are important theoretical problems in CS and so much more potential to be unlocked in other fields s.t. the core study remains necessary to understand how it can be better applied.

Also, there are areas in computer science that wouldn't be 100% applicable to, say, bioinformatics (e.g. cryptography) but may be very applicable in a much more generalized sense. I don't think we've delved into computer science for long enough that its study can be relegated and subjugated within other fields. That it is taught alongside courses in other fields speaks more to its practicality than its diminishing relevance as a standalone field.

I think there is a danger in Computer Science departments trying to become more theory oriented in order to survive. If applications are moving to their own departments in Business and Bioinformatics, CS may decide to become more theoretical. If this happens, though, growth in terms of research money will become even harder to find.

I've heard that discussed at some CS conferences as the "mathematics trajectory", which a lot of computer scientists think is best avoided. Essentially, becoming a pure discipline with little funding (because the applications work has gone elsewhere), and with a position in the university that's largely internally funded by your professors teaching intro-level "service" courses for other departments, the way math departments get a lot of their teaching credits by teaching Calc 101 and similar.

I don’t see the problem here as long as we really are talking about computer science. The University of Cambridge has two departments within the mathematics faculty already, one studying pure maths and statistics, the other applied maths and theoretical physics. Separating computer science research and teaching into a third element doesn’t seem inconsistent with that approach to me (though that’s not how the Computer Laboratory in Cambridge actually became what it is today).

Now, if we’re talking about teaching people to program simple software tools, as a useful practical skill like modelling something using calculus or writing a clear report, that’s a different question. There is room in the world both for specialists who are experts in a field and study the hard problems and for practitioners who aren’t experts but know enough to get useful things done. A full computer science course in a dedicated department is not a good fit for the latter group, who really need solid basic programming skills and perhaps a little software engineering knowledge.

With respect, I think the author confuses the roles of specialization and generalization. There are general skills within the domain of Computer Science that can be applied to other fields. Just like there are general skills from Statistics that apply across other fields, or general skills from English, or general skills from various disciplines of Mathematics. One can apply these skills across disciplines without needing to specialize, per se, in the discipline from which the skill derives.

Alternatively, one can decide that the skill's parent discipline is worth specializing in. If I'm a Biology major, and I discover that I am getting a kick out of the Statistics skills I employ in my major, even more so than I enjoy my major, I can decide to major in Statistics instead. This does not obviate the need for the two fields as separate fields of study. Nor does it present any real indication that one field is going to be collapsed into the other. The sets of Biology and Statistics have many intersections, but they are not the same set.

And that's the key: the fields are intersecting, and occasionally even overlapping, but they're not entirely so.

At the risk of sounding even nerdier, maybe this concept comes more naturally to those of us who've played an RPG or two in our day. Think Skyrim, if you've played it. You can build a melee fighter who draws a few skills from the magic skill tree. Or you can be a pure mage. Or a pure fighter. Or what have you. At the same time, there are only so many skill points you're allotted -- so you can't pick all of each tree. It works the same way in life, really. It's probably better to be a master of one trade with a few skills from the others, than to be a jack of all trades and master of none.

This has been my daily admission of dorkitude. Thanks for listening.

While i agree with your point, i think the article(as a whole) doesn't really contradict with your argument. It does presume(nay project) that at some point in the timeline of mankind CS departments will perish. but my imagination tells me that would not be because people don't study it, but rather because there is not enough funding(as dictated by the economy) to support a separate dept. for computer science. Note that doesn't say anything about the scope of work, or scope for the subject to revolutionize the world/science as we(mankind/humanity/perhaps even transhumans :-P ) know at that instant in the timeline. So unless you think it follows from your argument that CS, will never reach a position to be considered by a majority of population to be undeserving a separate department, i don't see a conflict at all. That's all i have to show off my nerdiness..:-P

I think that's fair, and well said. But I guess I just don't think that day will come. There will always be people who want to specialize in CS -- in the actual science of computers, and not just coding -- and these people will be different from the people who want to code, but don't want to get too deep into the weeds in the theory. Perhaps we'll see a day where CS and programming are separated out, and CS becomes a purely theoretical discipline, while programming is more like "Applied" CS. It's my understanding that this is already happening at some schools. But I don't think CS is in danger of disappearing as a distinct discipline.

I could be wrong, obviously, and I am not steeped heavily enough in the politics and ways of the world of academia to know for sure how these things go. But the fundamental discipline seems pretty distinct from others, even others like it.

In undergrad, for instance, I had the great fortune to take a CS class from Dr. David Gelernter. We didn't touch a single line of code in that entire semester, but I learned more about the fundamentals of computers than I have in any other class. Conversely, I had friends in the class who were bored to tears, or frustrated to kingdom come, by the subject matter -- who found the whole class impractical and not immediately useful. The phrase "intellectual masturbation" was tossed around pretty frequently. But honestly, different strokes for different folks. And that fact will keep CS alive and distinct.

Is there a "science" of computers? Not sure its science, maybe it is Maths? Or a part of Physics?

I might imagine one day C.S. departments might merge back into the Maths department, where (IMHO) C.S. most properly belongs.

This is like saying Chemistry should be merged back in with Physics as one is just a deep study of the application of the other. The edges between related departments are always going to be blurred, but it's still practical to have things separated out into different departments. I think that we'll just continue to so lots of faculty with joint or honorary positions across several departments. Hopefully this will also encourage students to collaborate in a more interdisciplinary way, too.

I could see theoretical CS merging back into math, and software engineering being part of engineering. The latter is more likely than the former.

I don't see anything merging back. The trend towards more and more specialization and compartmentalization has been going on forever and there's no reason to believe it will be inverted any time soon.

This sentiment comes up often enough that I wrote a brief essay on it. HN submission with discussion: http://news.ycombinator.com/item?id=3928276

My alma mater (The Ohio State University) has multiple Computer Science programs. There is an Associate program, an Arts & Sciences CS program, and a computer Science & Engineering program which is part of the Engineering school. I could see the Arts & Science program merging with the Math department, but the Engineering school will most likely hold onto the CS&E major.

There will be "Applied Computer Science" and "Pure Computer Science". I could see the latter going this way.

Why are math departments still around? My engineering professors insisted on teaching probability and linear algebra.

Gosh what a silly argument, claim that 'computer science' as a discipline will fade away based on the evidence that computer classes are taught in different disciplines.

That claim is simple to rebut with the example of mathematics, which are taught in various forms for nearly every department and yet there are still mathematics departments.

It would have been insightful to observe that basic computer science knowledge is becoming an essential part of any curriculum. That elevates it to the level of 'broadly applicable skill' like math, and composition. I believe its important as early as secondary school but those courses are still foundering for direction (I've seen 'intro to powerpoint' as a computer class which is more like 'typing' was than basic computer skills)

As long as there are computer science students.

Answer: for a long time to come. As long as there are students looking for credentials and certification in the form of degrees, there will be programs that cater to them.

You should see some of the outlandish courses and degrees offered by some schools.

As a bioinformatician, I've been asked all of the following questions by (postdoc and above) biologists:

- What is a binomial distribution?

- What is the difference between a probability and a p-value?

- If I have p-values for a case and control group, can I just subtract the log10 p-value for my control from my case group?

From fellow "bioinformaticians", research scientist level and above, I've been asked:

- How can I load this MySQL dump you sent me into my database? phpMyAdmin doesn't read MySQL dumps.

- What is a likelihood ratio (from a PI!!)

So, needless to say, I feel that I have some good job security.

Does this writer assume CS is simply a tool used in other industries? There should be CS departments as long as we want to advance the power and reach of computing.

How long will there be physics departments? Physics is just theoretical engineering and chemistry. Most actual users of physics are engineers or chemists; why would anyone need a separate physics department?