topics: |  linguistics | morphology

Excerpts

we have to make a distinction between the notion ‘word’ in an abstract
sense (LEXEME) and the notion ‘word’ in the sense of ‘concrete word as used
in a sentence’.  The concrete words walk, walks, walked, and walking can be
qualified as WORD FORMS of the lexeme walk. ... p.3

The words walk, walks, walked, and walking show a relationship in form and
meaning of a systematic nature, since similar patterns occur for thousands of
a% other verbs of English. The subdiscipline of linguistics that deals with such
patterns is called morphology. The existence of such patterns also implies
that word may have an internal constituent structure. For instance, walking
can be divided into the constituents walk and -ing. Therefore, morphology
deals with the internal constituent structure of words as well.

forms such as walking are formed from the lexeme walk according to rules, and
therefore need not be specified individually in the dictionary.  For example,
my English–Dutch dictionary does not mention the adverbs correctly and
economically (it only gives the adjs correct and economical). On the
other hand, it does specify the adverb hardly....  Is there a principled
reason? ... the meaning of hardly cannot be predicted from that of hard and
-ly. p.4

Morphological Parsing

when searching for information on tax issues, you would not like your search
engine to retrieve documents with the words taxi, taxis, taxon, or taxonomy
that also begin with the letter sequence tax.  This example shows that
analysis of the systematicity in the relations between words is essential for
the computational handling of language data. What we need for this purpose is
a morphological PARSER, a computer program that decomposes words into
relevant constituents: tax-ation, taxable, and tax-abil-ity.

taxation, taxable, and taxability : WORD FAMILY. 

[taxability: -ity can be added only after tax+able is available, so it is a
two step process, given by the bracketing structure  (p.10)
	[ [ [tax]N [-able]Adj] [-ity]N ]

What is a word?

Determining if a particular linguistic unit is a word is not always easy,
certainly for languages without a written tradition. Even for English we
might not be certain. Why is income tax to be considered as a word rather
than a phrase? After all, its constituents are separated by a space in its
spelt form. p.5

Word-formation is traditionally divided into two kinds: DERIVATION and
COMPOUNDING. Whereas in compounding the constituents of a word are
themselves lexemes, this is not the case in derivation. For instance, -ity
is not a lexeme, and hence taxability is a case of derivation. The word
income tax, on the other hand, is a compound since both income and tax are
lexemes. Changing the word class of a word, as happened in the creation of
the verb to tax from the noun tax, is called CONVERSION, and may be
subsumed under derivation.

Many compounds - meaning may be inferrable from the constituents -
e.g. "bottle factory". 

[This is doubtful.  Surely a factory could be shaped like a bottle, or be
housed in one even, etc. ] 

Systematicity

	buyer = buy + er

buy is a FREE MORPHEME or LEXICAL MORPHEME 
	[or SIMPLEX WORD, one that cannot be decompose further]
-er : AFFIX or BOUND MORPHEME
buyer : COMPLEX WORD p.7

e.g. eater painter seller sender: 
	[ [x]V [-er]N ] = one who x's 

but not all meaning-change patterns are morphemic - e.g. ear -> hear
is not systematic - h- does not appear in any other usage. 

even systematic forms may not indicate a morphoology, e.g. many fish names
in dutch end in -ing:
	bokking “bloater”, 
	haring 	“herring”, 
	paling 	“eel”, 
	wijting “whiting”

but this is not a morphological segmentation since the roots (which exist)
have completely unrelated meanings.

[English: raspberry, strawberry etc.?  also there is "cranberry". ]

Reduplication

examples (Uhlenbeck 1978: 90) from Javanese:

(13) a. full reduplication:
	baita “ship” baita-baita “various ships”
	səsupe “ring” səsupe-səsupe “various rings”
	omaha “house” omaha-omaha “various houses”
     b. partial reduplication:
	gəni “fire” gəgəni “to warm oneself by the fire”
	jawah “rain” jəjawah “to play in the rain”
	tamu “guest” tətamu “to visit”

In partial reduplication, the prefix copies the
first consonant of the base followed by schwa [ə]. p.35