Ph.D. Qualifying Examination for Linguistics—Saudade Gezellig Collateral Descendant of Lingua Pranca Contents Handy Definitions for Newcomers to the Field of Linguistics—Ken Miner and David J. Peterson

On the Two Main Types of Computational Linguists

J. Random Linguist

In years gone by, linguists and philosophers displayed a great deal of interests in each other’s work. Nowadays, not so much, at least on the linguists’ side. I have no idea what philosophers are doing these days. In any event, computer scientists and linguists have begun to intermingle to a considerable degree. Linguists can be found attending computer science courses in significant numbers and several computer scientists have at least heard of linguistics. Where the two are officially conjoined, natural language processing results (occasionally to a statistically significant degree). But many computational linguists are really either just computer scientists who know what a parser is (even if they think “real” ones are best built with lex and yacc), or just linguists who know more about computers than their thesis advisors. Nonetheless, it has been suggested that it can be difficult to tell a linguist from a computer scientist these days. For the benefit of anyone who is interested in drawing that distinction, I submit the following compendium of spot checks:

The linguist is the one       The computer scientist is the one
 
who reads J.R.R. Tolkien to learn Quenya and get ideas for their next conlang who reads J.R.R. Tolkien to learn about all the creatures in Moria and to get ideas for their next LARP
 
who uses example sentences that contain subordinate clauses who uses example diagrams that are never explained or even referred to in the text
 
who talks about “daughter” and “sister” nodes in trees who talks about “child” and “sibling” nodes in trees
 
who goes to great length to formalize the idea that lines in their syntax trees can’t cross who never considered foolishly crossing lines in their binary trees (and really only use hash tables these days)
 
who carefully and formally defines C-command, thus confusing another generation of Syntax I students who blithely says “over one and down into the sub-trees”, thus confusing another generation of CS 101 students
 
who prattles on at dinner about how cow and beef are etymologically descended from the same PIE root who owns and regularly eats with a laser-cut titanium spork
 
who owns an electronic copy of the OED, but hasn’t figured out how to search it properly, and so has only looked at entries starting with “aa-” and “ab-” who owns a paper dictionary of English but only looks up words online (and then only for other people)
 
whose IQ drops 30 points when writing a textbook whose ability to write coherent prose drops by 50% when writing a textbook
 
who can play chess who can write a program to play chess tic-tac-toe
 
who knows the 7-times table who can write a one-line Perl program in 45 seconds to generate the 7-times table when needed
 
who uses Big-O complexity notation in a paper on Mongolian vowel harmony who refers to Chomsky in a paper on optimization of red-black trees
 
who knows the technical distinction between slang and jargon who knows how Jargon and the author’s pseudonym are related

Ph.D. Qualifying Examination for Linguistics—Saudade Gezellig
Handy Definitions for Newcomers to the Field of Linguistics—Ken Miner and David J. Peterson
Collateral Descendant of Lingua Pranca Contents