First of all learning Latin had a definite use. In an era where every educated person learned Latin, not knowing it shut you out of many areas. And even today if you need to access historical records, it is critical to know Latin. But as I said elsewhere, saying that a subject has a use doesn't mean that it is useful enough that everyone should go out and learn it. Quite the contrary, because of constraints on time we do not learn many useful subjects, and should choose wisely.

It also doesn't mean that your examples are well-chosen.

Law I cannot comment on, knowing Latin may be useful there. But people I have known in various sciences and medicine have told me point blank that learning Latin is not very useful for what they do. Yes, they use Latin-derived words. But most of what they have to know isn't those words, and learning the ones that they need to know is such a small part of what they do that it would make no sense to learn Latin first.

So I have to decide whether to believe a random stranger on the net making sweeping claims about various professions versus accepting what PhDs and MDs have told me about what they do. I would that all my decisions were that easy!

Going further, what you say about various Romance languages (and about English - which got a lot of vocabulary from French in particular) is definitely true. However it is a disingenuous argument since learning any Romance language for the same effort both teaches you a language used by millions alive today and gives you the same head start on other related languages. Which makes Latin an inefficient path to that result.

And that is why Latin is no longer a basic part of our general education. Most of the reading and writing people do is in their own vernacular language(s), scholarly discourse takes place almost exclusively in them, and even Christian worship is no longer generally conducted in Latin (a process that took from Martin Luther to the late 1960's). Given this, learning Latin simply doesn't make very much sense for most people. It held on for a long time based on inertia and general acceptance that somehow someone who didn't know Latin wasn't really educated. Learning Latin taught, so the argument went, "analytic thought". Those declensions had to be good for something!

That proved not to be enough. So eventually Latin was discarded, more useful topics replaced it, and life went on. As with the loss of Euclid's Elements as a basic textbook a few decades earlier, something was lost and on the whole rather more was gained.

PS I know that spelling flames are silly, but I have to admit to finding it ironic when someone who is telling me the importance of developing a better vocabulary gets the spelling of "medicine" wrong...

Well-rounded developers should be able to document their work and communicate with others. That requires a decent vocabulary, something that learning Latin definitely helps with.

yes, people who use calculus in PERL scripts are the same people who would document their code in Latin.

Congratulations, you've completely missed the point. We should teach CS majors calculus not so they can write numeric-integration code (which they'd be better off getting from Numerical Recipes instead), but so they learn how to analyze and solve problems in the context of formal rule sets. Programming fits that description.

I've taught undergraduate programming labs for the past three years: very concrete, applied stuff (the course is an introduction to software engineering; we teach them how to write programs, not how to write code that compiles). The classes are very eclectic: many CS majors (some interested in theory, some in practice), many engineers, and a strong minority of other Science, Arts and Business majors. The students who've taken courses involving more abstract formal reasoning (Math, Philosophy, Physics, etc) are consistently better able to understand the concepts and do much better on the coursework than the ones who focus on nuts and bolts. The best student I've had in that course was an English major with a minor in Philosophy.

people with CIS degrees are the people working on help desks and doing data entry.

Then what are they doing at a university?

for a BA in CS why not skip the higher math and try to teach them how to debug and optimise code.

You can't teach someone how to debug or optimize code. (Glub knows I've tried.) The best you can do is teach them how to think logically and abstractly about the code, which involves teaching them how to think logically and abstractly.

Math tends to be better at that than software engineering.

(CS theory/algorithmics is also good, but that's just another math course, really.)

so that everything that they write for the first couple of years after getting their degree doesn't have to be re- written by someone with more experience.

The problem, in my experience teaching students and in the private sector, isn't that people know calculus, but not how to code: it's that they can't think abstractly about the code. They can catch syntax errors, but can't solve logic errors to save their lives (or jobs). With a very few exceptions, the people whose code has to be rewritten are the same ones whose reply to required math courses was "when am I ever going to have to use this?"

--
The hell with paco, vote for Erudil!
:wq