Corpora: "language engineering"

[Doug Cooper]

"Language engineering" (nee "text hacking" circa late 1970's, then
briefly "natural language understanding") is none other than natural
language processing, engaged in a brave attempt at regaining its
respectability after an lengthy stint as the doxy of AI.  The breakdown
of the relationship may be traced to the long public humiliation NLP
endured during the "AI gap" era of the mid 1980's (the heyday of the
Japanese Fifth Generation computing project and the US Strategic
Computing Initiative).

    The term "engineering" is just one element of an extended metaphor --
which includes terms like "tools" and "workbench" -- used to add a sense
of vigor and brawn to work in this area.  The alternative "computational
linguistics," although common within the field as an appropriate description
of much of the logical evolution of text hacking, is apparently not quite
sexy
enough for universal acceptance -- possibly because the pause requred to
explain the meanings of both "computational" and "linguistics" to potential
supporters tends to kill any mood conducive to providing funding.

   In the above sense, the word "engineering" should be recognized as
a marker of social register, largely bleached of its original semantics.  It
enjoyed a vogue in the 1970's as a means of raising the social status of
one's employment , but was overused to such an extent that the phrase
 'X engineering' or  'I'm a Y engineer' became a guaranteed laugh line,
and the term lost favor in polite company, much like the extended pinky
of an earlier era.  It occasionally reappears in the research findings of the
cosmetics industry (eg. 'rehydration engineering'), as reported in many
fashion industry journals, and regularly summarized in The Journal of
Irreproducible Results.

  A parallel usage, as an adjective meaning 'imposed from above in
accordance with firm principles of pseudo-science,' became popular at
roughly the same time, particularly in the phrase 'social engineering.'
It is
almost invariably found in close collocation with such terms as "attempts at"
and "failed."  Many researchers have pointed out the unintended irony
found in the frequent pejorative use of "social engineering" in papers
whose letterheads include the (apparently non-judgemental) term "social
science."

   In closing, I should point out that the field of text hacking has
undergone
sweeping changes in the past 25 years, as evidenced by the replacement
of awk, grep, and sed with Perl, and the universal substitution of the hash
for
the associative array.  Surely, the great historical figures (Kernighan,
Lesk,
Cherry, et al) would hardly know what to make of our modern science.

  Yours in discourse,
  Doug Cooper
__________________________________________________
          1425 VP Tower, 21/45 Soi Chawakun
        Rangnam Road, Rajthevi, Bangkok, 10400
    doug at th.net (662) 246-8946  fax (662) 246-8789

  Southeast Asian Software Research Center, Bangkok
  http://seasrc.th.net         -->  SEASRC Web site
  http://seasrc.th.net/sealang --> SEALANG Web site