Transcript Ling 001, Week 4

Ling 001
Word Structure, Part II
Outline
• In looking at morphology we are examining the
relationships that words have to one another, and to
the morphemes that are assembled into complex
structures
• Two further themes
– Different kinds of morphology: inflection vs. derivation
– What is a word? What can go into a word?
• We’ll see along the way how languages differ in
terms of the distinctions we’ve introduced
• Conclude with questions about morphology and
syntax
I. Inflection and Derivation
• Inflection: Creates new forms of the same
word in a way that introduces or expresses
different grammatical properties, while
retaining some core notions of meaning (and
category)
• Example:
Play and Played describe the same action, but
situate it differently in time.
Inflectional categories
• Languages differ with respect to which categories are
expressed inflectionally on e.g. verbs. English, for
instance, expresses Person (1st person, 2nd person,
3rd) in a limited way, as well as tense:
Present
Past
1s
praise
prais-ed
2s
praise
prais-ed
3s
praise-s
prais-ed
1p
praise
prais-ed
2p
praise
prais-ed
3p
praise
prais-ed
(s = singular, p = plural)
• Notice that marking for Person is not found in Past
Comparison
• The expression of such inflectional categories
is limited in English. Compare Latin (lauda:re
‘praise’):
Present
Past (imperfect)
1s
laud-o:
lauda:-ba-m
2s
lauda:-s
lauda:-ba:-s
3s
lauda-t
lauda:-ba-t
1p lauda:-mus
lauda:-ba:-mus
2p lauda:-tis
lauda:-ba:-tis
3p lauda-nt
lauda:-ba-nt
Comparison, cont.
• In the English and Latin comparison, we are
talking about the same abstract categories in
some sense: Tense and Person/Number
• Languages express different notions in
verbal marking:
• Classical Greek: Dual as well:
• Lu-ei `he/she/it looses’
• Lue-ton `they-2 loose’
• Luo-usi `they loose’
Another example
• Example 2 : (Some) verbs in Tepetotula Chinantec
differ for whether the object is animate or inanimate:
The verb here is ‘abandon’:
Inanimate
Animate
1s
1p
2(s/p)
3(s/p)
tíLM
tíLM
tíLM?
tíM
téNLM
téNLM
téNLM?
téNM
So, if you want to say ‘I abandoned my friend’ versus ‘I
abandoned the house’, you have to use different verb forms
Synopsis
• Languages differ in terms of
– What type of information is expressed in different
categories of words; and
– How many distinct means of marking such
differences there are
• A further point of cross-linguistic difference
concerns how much can fit into a single word,
and how we are going to define word for
different languages in the first place (see
below)
Inflection, cont.
• Some general properties associated with inflection,
generalizations which hold for the most part:
– Inflection does not change syntactic categories. E.g. kick-s is
still a verb, even with its inflectional suffix
– Inflection expresses grammatically required features or
relations (e.g. agreement, tense, etc.)
– Inflectional morphemes occur outside of derivational (see
below) morphemes: ration-al-iz-ation-s
• As a general way of thinking of this, inflection creates
new forms of the same word; derivation is thought to
create a ‘different’ (but related) word
• Some inflectional morphemes in English:
--ed (past tense), -s (plural), etc.
Derivation
• As a basic working definition, derivational
morphology creates new words from existing ones.
Basic properties:
– Change of category or part of speech (noun, verb, adjective)
is possible: pay, pay-ment
– New meaning added: e.g. re-do means to ‘do again’
– Inflection often has syntactic connections outside of the
word, (e.g. agreement relates a subject to a verb). This is not
so if we have e.g. kind/unkind; the change doesn’t relate to
anything external
– Sometimes not productive (it sometimes doesn’t attach to
some words) or unpredictable meanings:
• Destroy/destruction; employ/*empluction/employment
• Transmit ‘send’; transmis-sion ‘sending’; ‘car part’
Derivation: Examples
Morpheme
Function
-(a)tion
verb --> noun
deviate, devia-tion
-al
noun --> adjective
institution, institution-al
-ize
noun --> verb
color, color-ize
-like
noun--> adjective
dog, dog-like
un-Karl Farbman-like
Further aspects of derivation
• Derivation is not necessarily categorychanging; sometimes it creates a new word
with the same category as the root/stem, but
with a different meaning:
king, king-dom
star, star-dom
• But nounhood is a property of -dom in this
case, as is clear from instances in which it
attaches to other categories:
free, free-dom
Some unpredictability
• In many cases, the same kind of derivational pattern
shows differences in form; take e.g. verb --> noun:
1) -al
refuse
refus-al
arrive
arriv-al
2) -ion
confuse
confus-ion
extend
extens-ion
3) -ation
derive
derivation
confirm
confirm-ation
4) -ment
confine
confine-ment
treat
treat-ment
This is in a sense allomorphy: the form of the nominalizing affix
is something that depends on what host the affix is attached
to (put differently, the different affixes only attach to certain
hosts)
Additional Interactions
• Often the distinction between derivation and
inflection is used as a helpful tool, not an
absolute distinction
• Consider some additional cases in terms of
our criteria above:
Formation of gerunds in -ing:
John destroyed the house.
John’s destroying the house (upset me).
Gerunds, cont.
•
Formation of the nominalization in -ing is
–
General: we can take whatever verbs we think of and form
such nominals
–
Shows no allomorphy: all such nominals show -ing.
Sometimes there is more than one denominal verb:
1) John’s destroying the city
2) John’s destruction of the city
There is a sense in which the second is more ‘nounlike’
than the first
•
General point: This type of case meets some of the
criteria for both inflection (regularity, productivity)
and for derivation (category change)
II. What’s in a word?
• Recall our division of morphemes along two
lines: free vs. bound and content vs. function:
Bound
Free
Content
crandog
Function
-ed
the
• Languages differ in terms of how they divide
up this cross-classification; many languages
have more morphological (bound) marking
than e.g. English
– Relatedly, languages differ in terms of what can go
in a ‘word’ (we can try to define word below)
Words
• One way to think of this is in terms of some
counting exercises; how many words in
John ate the apple
• How about
I’ll eat the apples later.
I will eat the apples later.
I didn’t eat any apples yesterday
Distinctions
• Phonological Words: An object that forms a
single unit for the purpose of phonology
• (Syntactic) Word: A single object for the
purposes of the syntax
Example:
I’ll eat the apples later.
Here I’ll is a single phonological word. But if we
think that this sentence has the same syntax
as I will eat the apples later, this single
phonological word is composed of two
syntactic words
Complex words
• Languages differ greatly in terms of what they
package into their words (relatedly, in terms
of what is expressed as bound or free)
• Some languages pack a great deal into single
(phonological) words:
English:
They treated us in that way
Hupa (California, Athabascan)
‘a:yanohch’ilah
Analysis
• The Hupa example:
‘a:yanohch’ilah
‘a-
ya- noh-
ch’i-
lah
thus PL 1Pl-Obj 3rdPl-Subj treat
• In this language and many others, what is
expressed in English with many free
morphemes is expressed in a single
phonological word, with many bound
morphemes
Incorporation
• Noun Incorporation
– Mapudungun
Ni chao kintu-waka-le-y
my father seek-cow-TNS-3s
‘my father is looking for the cows’
• Here, the meaning of the phrase “look for cows” is expressed in
a single word (they can express it with a separate noun as well).
• This is similar in many ways to what happens in compounding in
English; remember truck driver. In English, though we can’t use
this as a verb *I truck-drive.
• But, in Mapudungun (and many other languages!) these
strutures are not restricted to nouns; they happen with verbs as
well.
‘How much’ Morphology
• Languages are often described in terms of whether
they have little (English, Chinese) or rich (e.g. Hupa,
Latin) morphological systems
• Further distinctions: whether meanings are
“combined” in morphemes, or separated into different
morphemes:
– English: from our islands
– Latin: insul-i:s
nostr-i:s
island-ABL.PL. our-ABL.PL
– Turkish: ada-lar-ImIz-dan
island-PL-OUR-ABL
Syntax/Morphology
• What do the examples on the last slide show?
At some level of description, languages
express the same meanings in different ways,
ranging from “more syntactic” (English) to
“more morphological” (Turkish)
• This suggests that there is no sharp dividing
line between a “word system” (morphology)
and a system for assembling words into
phrases etc. (syntax)
• Some more thoughts along these lines…
Morphology and Syntax, cont.
• With morphology we refer to the study of
words and their structure, while with syntax
we refer to the structure of larger objects
(phrases, clauses)
• Examples:
– The black board (phrase = syntax)
– The blackboard (2nd part is a word=morphology)
• In some cases, the distinction between these
two domains of study is blurred as well
Interactions between syntax and
morphology
• Consider how comparatives and superlatives work in
English
– Comparative: tall, tall-er
• In cases of this type, the comparative seems to be a
kind of (inflectional?) morpheme, creating a
comparative adjective from an adjective
• But:
Think of more adjectives
smart, smart-er intelligent, *intelligenter
Note that the comparative of intelligent requires a
phrase:
more intelligent
More examples
• A more difficult case occurs with English in the
phenomenon called do-support
• Consider a normal past tense sentence:
John play-ed football yesterday.
Notice that the (bold-faced) past tense morpheme -ed
appears on the verb play
Now the negative equivalent:
John di-d not play football yesterday.
Here we see past on do, in did, which is the past tense of that verb.
The past tense, which appears as part of the word in the first
example, occurs in a different word in the second example
• A consistent treatment of these facts involves a structure in
which the tense morpheme -ed occupies a different syntactic
position from that occupied by the verb