🚀 go-pugleaf

I promised in a couple of messages that I would define some terms that hereto-
fore I had used as semantic primitives.  I intended to do this in a response to
several messages from Mark Hubey, but that is growing into a long project, and
the definitions need to be done anyway.

regular:  Predictable by historical or comparative expectation.

similar:  Showing like patterns of form and function when compared.

paradigm:  An ordered arrangement of forms (most frequently words, but not of
	necessity) intended to show in compact form the differences and
	similarities (see "similar" above) thereof.  Canonical examples:  The
	forms of the present active indicative of Latin verbs, or the forms of
	the noun in all 15 cases in Finnish.

	It has nothing to do with any misuse of the term by philosophers of
	science.

relatedness:  In linguistics, a *defined* term, with a meaning only with
	respect to the history of languages.  To speak of languages as related
	*entails*--not *implies* but *entails*--a claim about their history.

	There is no need to bring in pseudo-sciences such as astrology or
	economics.  There is no "utility function" to decide whether a group of
	languages are related or not.  This is strictly an historical issue.

	Now, there *is* a methodology for testing a hypothesized relationship:
	The comparative method.  This has not been shown to be insufficient for
	the task, although it has been *claimed* to be insufficient by some
	whose hypotheses have not been borne out by it.

	Some objection has been made to the use of "genetic" terms in the
	description of relationships among languages, since "mammals have two
	parents."  I would point out that there are other models in biology
	besides the mammals, even if that were the source of the original
	analogy.  For instance, amoebae are asexual producers of offspring,
	disappearing in the process; they mutate from time to time; and they
	exchange material from time to time.

	The model, however, originates with scholars in the Renaissance, who
	worked to reconstruct original texts of documents transmitted from the
	Classical period via monks who did not know (well) the languages of the
	texts.  These philologists built family trees of manuscripts on the
	model of the lines of succession of European royal houses, where only
	rarely was more than one parent important in terms of inheritance.

comparison:  A methodology for determining whether a group of languages are
	related or not.  It involves the painstaking collection of large
	amounts of vocabulary from the languages in question, including as many
	paradigms (see above) as possible.  These paradigms are then compared
	to each other in order to find similarities (see "similar" above) first
	of morphological or syntactic pattern.  If similarities are found, the
	morphemes in question are re-examined to determine whether regular (see
	above) correspondences in their phonological makeup can be found.

	For example if some languages show only velar stops while others show
	velar stops before back vowels and palatal fricatives before front
	vowels, the comparison is allowed to to conclude that some change has
	occured.

	If there is partial regularity, such as a usual correspondence between
	voiceless stops in some languages and voiceless fricatives in others,
	but in certain cases between the same voiceless stops and *voiced*
	fricatives in the same other languages, the comparison is allowed to
	conclude that some *conditional* change has occurred, for which the
	conditioning may not be immediately apparent.

linguistics:  A group of studies focusing on the description of various aspects
	of language, with an eye towards understanding them.  These studies are
	of several types, and are joined under a single rubric by their focus,
	much as the various branches of physics are all considered to be
	"physics."

	There are a number of ways to divvy up the field of linguistics.  One
	major division is that between synchronic and diachronic studies
	(Saussure 1916).  This is a distinction between the study of speech in
	a single time-frame versus the study of speech across time-frames.

	Across this dichotomy is a breakdown by special area of interest:
	Phonetics (acoustic study of spoken language; study of perception of
	spoken language; study of production of speech); phonology (study of
	synchronic pattern of occurrence of speech sounds; diachronic study of
	change between one synchronic pattern of speech sounds and another);
	morphology (study of synchronic occurrence of patterns of groups of
	phonemes and their conjoining; diachronic study of change between one
	synchronic pattern of such phoneme sequences and another); syntax
	(study of patterns of use of morphemes to make up utterances;
	diachronic study of change from one such pattern to another); semantics
	(study of "meaning"--the definition of which is one research topic in
	this subfield); pragmatics (study of situation external to an actual
	utterance which contributes to the "meaning" thereof--the least
	susceptible to diachronic study, in my opinion).

	The synchronic side is generally referred to as "theoretical" linguis-
	tics, although there are theories on the diachronic side as well.  This
	group of studies is susceptible of investigation by experimentation,
	either instrumented (as, for example, acoustic phonetics) or introspec-
	tive (as, for example, acceptability judgments in syntax).

	The diachronic side--the study of languages in a temporal dimension--
	uses the results garnered by "theoretical" linguistics, but in addition
	uses methodologies similar to those of other *historical* studies,
	methodologies which simply do not admit of mathematical manipulation.

I hereby release this message in its entirety for use in the list of Frequently
Asked Questions, should the keepers of that worthy document wish to do so.
-- 
Rich Alderson   'I wish life was not so short,' he thought.  'Languages take
                such a time, and so do all the things one wants to know about.'
                                                        --J. R. R. Tolkien,
alderson@leland.stanford.edu                              _The Lost Road_