Sunday, October 7, 2012

Morphology


Morphology

1.1       What is Morphology?
The term morphology is generally attributed to the German poet, novelist, playwright, and  philosopher Johann  Wolfgang von  Goethe  (1749–1832), who  coined  it early  in the nineteenth century in a biological  context.  Its etymology is Greek:  morph- means ‘shape,  form’,  and  morphology is the study of form  or forms.  In biology  morphology refers  to the  study of the form  and  structure of organisms, and  in geology  it refers  to the study of the  configuration and  evolution of land  forms.  In linguistics morphology refers  to the mental system involved in word formation or to the branch of linguistics that  deals  with  words, their  internal structure, and  how they  are formed.
1.2       Morphemes
A major  way  in  which  morphologists investigate words, their  internal structure, and how they are formed is through the identification and study of morphemes, often defined as the smallest linguistic pieces with a gram- matical  function. This definition is not meant to include all morphemes, but it is the usual  one and a good starting point.  A morpheme may consist of a word, such  as hand, or a meaningful piece of a word, such  as the -ed of looked, that  cannot  be divided into  smaller meaningful parts.  Another way in which morphemes have been defined is as a pairing between sound and  meaning. We have purposely chosen  not to use this definition. Some morphemes have no concrete form or no continuous form, as we will see, and  some  do not have  meanings in the conventional sense  of the term.
You may  also run  across  the term  morph. The term  ‘morph’  is some- times   used   to  refer   specifically  to  the  phonological  realization  of  a morpheme. For example, the English  past  tense  morpheme that  we spell
-ed has various morphs. It is realized as [t] after  the voiceless  [p] of jump (cf. jumped ), as [d] after the voiced  [l] of repel (cf. repelled ), and as [@d] after the  voiceless  [t] of root or the  voiced  [d] of wed (cf. rooted and  wedded). We can also call these  morphs allomorphs or variants. The appearance of one morph over another in this case is determined by voicing  and  the place  of articulation of the final  consonant of the verb  stem.
Now   consider the  word reconsideration. We  can  break   it  into  three morphemes: re-, consider, and  -ation. Consider is called the stem. A stem is a base morpheme to which  another morphological piece is attached. The stem can be simple, made up of only one part, or complex, itself made up of more  than  one piece. Here  it is best to consider consider a simple  stem. Although it consists  historically of more  than  one part,  most  present-day speakers would treat it as an unanalyzable form. We could also call consider the root. A root is like a stem in constituting the core of the word to which other  pieces  attach,  but  the  term  refers  only  to morphologically simple units. For example, disagree is the stem of disagreement, because it is the base to which  -ment attaches, but agree is the root. Taking  disagree now, agree is both  the stem  to which  dis- attaches and  the root  of the entire  word.
Returning now  to reconsideration, re- and  -ation are both  affixes,  which means that  they  are  attached to the  stem.  Affixes  like re- that  go before the stem  are prefixes,  and  those  like -ation that  go after  are suffixes.
Some  readers may  wonder why   we  have   not  broken -ation  down further into  two  pieces,   -ate and   -ion,  which   function independently elsewhere. In this  particular word they  do  not  do  so (cf. *reconsiderate), and  hence  we treat  -ation as a single  morpheme.
It is important to  take  very  seriously the  idea  that  the  grammatical function of a morpheme, which  may  include its meaning, must  be con- stant.  Consider the English  words lovely and  quickly. They both  end  with the suffix -ly. But is it the same  in both  words? No – when we add  -ly to the adjective quick, we create  an adverb that  describes how  fast someone does  something. But  when we  add  -ly to  the  noun  love, we  create  an adjective. What  on  the  surface  appears to be a single  morpheme turns out  to be two.  One  attaches to adjectives and  creates  adverbs; the  other attaches to nouns and  creates  adjectives.
There  are  two  other  sorts  of affixes  that  you  will  encounter, infixes and  circumfixes.  Both are classic challenges to the notion of morpheme. Infixes  are segmental strings that  do not  attach  to the front  or back of a word, but  rather somewhere in  the  middle. The  Tagalog  infix  -um-  is illustrated below  ( McCarthy and   Prince  1993: 101–5; French   1988).  It creates  an agent  from  a verb  stem  and  appears before  the first  vowel  of the word:

(1)   Root        -um-
/sulat/ /s-um-ulat/      ‘one who  wrote’
/gradwet/   /gr-um-adwet/  ‘one who  graduated’

The existence  of infixes  challenges the traditional notion of a morpheme as an indivisible unit.  We want  to call the stem sulat ‘write’ a morpheme, and yet the infix -um- breaks  it up. Yet this seems to be a property of -um- rather than  one  of sulat. Our  definition of morphemes as  the  smallest linguistic pieces  with  a grammatical function survives this challenge.
Circumfixes are  affixes  that  come  in  two  parts.  One  attaches to  the front  of the word, and  the other  to the back. Circumfixes are controver- sial because it is possible to analyze them  as consisting of a prefix  and  a suffix  that  apply to a stem  simultaneously. One  example is Indonesian ke . . . -an. It  applies to  the  stem  besar ‘ big’  to  form  a  noun   ke-besar-an meaning ‘ bigness,  greatness’ ( MacDonald 1976: 63; Beard 1998: 62). Like infixes,  the  existence   of  circumfixes challenges the  traditional  notion of morpheme ( but  not  the  definition used  here)  because they  involve discontinuity.
We  will  not  go  any  more  deeply here  into  classical  problems with morphemes, but the reader who  would like to know  more  might consult Anderson (1992: 51– 6). 
 
1.3       Morphology in Action
We  would like  to  explore   the  idea   of  morphology more   deeply  by examining some  data.  These  are  examples of  morphology in  action  – morphological facts of everyday life.

1.3.1    Novel words and word play

If you  had  been  walking down the  street  in  Ithaca,  New  York,  a few years  ago, you might have  looked  up and  seen a sign for the music  store “Rebop,” a name  that  owes  its inspiration to the  jazz term  rebop.1 Rebop was originally one of the many  nonsense expressions that  jazz musicians threw into  their  vocal  improvisations, starting in the  early  1920s. In the
1940s, rebop became  interchangeable with  bebop, a term  of similar  origin, as  the   term   for  the   rhythmically  and   harmonically  eccentric   music played by young black musicians. By the 1950s the name  of this musical style was  quite  firmly  established as simply  bop.2 Today,  the original use of rebop is known only  to  cognoscenti, so  that  most  people who  pass by  the  store  will  be  likely  to  interpret the  word as  composed of  the word bop and  the  prefix  re-, which  means approximately ‘again’.  This prefix  can attach  only  to verbs,  so we must  interpret bop as a verb  here. Rebop must   therefore mean   ‘ bop  again’,  if  it  means anything  at  all. And  this  music  store,  appropriately,  specialized in  selling  used   CDs. There’s  something going  on  here  with  English  morphology. Of course, rebop is not  a perfectly well-formed English  word. The verb  bop means something like  ‘bounce’,  but  the  prefix  re- normally attaches only  to  a verb  whose meaning denotes an accomplishment. The verb  rebop there- fore  makes  little  sense.  But names of stores  and  products are  designed to  catch  the  consumer’s attention, not  necessarily to  make  sense,  and this  one does  so by exploiting people’s knowledge of English  in a fairly complex way  and  breaking the rules  so as to attract  attention, as verbal art often  does.
Consider now  the following phrases, taken  from  a Toni Braxton  song:
Unbreak my heart, uncry these tears.
We have never  seen anyone unbreak something, and  you certainly can’t uncry tears,  but  every  English  speaker can understand these  words. We all  know  what  it  means to  unbreak somebody’s heart  or  to  wish  that one’s heart  were  unbroken. If we asked  somebody, “unbreak my heart,” we  would be  asking  them  to  reverse the  process  of having our  heart broken. We can visualize “uncry these tears,”  too – we just think  of a film running backwards. We can understand these  words because we know the  meaning of un-,  which  basically  reverses or  undoes an  action.  The fact that  these  particular actions,  breaking a heart  and  crying  tears,  can- not be reversed only  adds poignancy to the song.
All human beings  have this capacity for generating and  understanding novel  words. Sometimes someone will  create  an  entirely new  word, as J. R. R. Tolkien  did  when he coined  the now-familiar term  hobbit (which, despite its popularity, is still not  listed  in the  2000 edition of the  Amer- ican Heritage Dictionary). But more  often  than  not,  we  build  new  words from  pre-existing pieces,  as with  unbreak and  uncry. We could  easily  go on to create  more  words on this pattern.
Novel  words are  all  around us.  Jerry  Seinfeld  has  talked about  the shushers, the  shushees, and  the  unshushables in  a movie  theater. Morley Safer was  dubbed quirkologist – expert  on  quirky people – on  a special episode of 60 Minutes. For those who hate buffets,  the TV character Frasier Crane  came  up  with  the  term  smorgsaphobia. Finally,  the  longest novel morphologically complex word we have  been able to find  on our own  in the daily  press  is deinstitutionalization, from  the New York Times.
These are everyday morphological facts, the kind  you run  across every day  as a literate speaker of English.  What  these  words – rebop, unbreak, uncry, hobbit, quirkologist, smorgsaphobia, and  deinstitutionalization  – have in common is their  newness. When  we see or hear  them,  they leap out at us,  for  the  simple  reason that  we  have  probably never  seen  or  heard them  before.  It  is  interesting that  novel  words do  this  to  us,  because novel  sentences do  not.  When  you  hear  a new  sentence, you  generally don’t realize  that this is the first time that you’ve  heard it. And  you don’t say  to yourself, “ What  a remarkable sentence,” unless  it happens to be one from  Proust or Joyce or some  other  verbal  artist.  Many  people have made the observation before  that  morphology differs  from  syntax  in this way.

1.3.2    Abstract morphological facts
Now  let’s move  to some more  abstract morphological facts. These are the kind  of morphological facts that  you don’t  notice  every  day.  They are so embedded in your  language that  you don’t  even think  about  them.  They are more  common than  the ones we have  just looked  at, but  at the same time  deeper and  more  complex.
If you  speak  English  and  are concerned about  your  health,  you  might say:

(2)   I eat one melon  a day.

Let’s imagine that  we  are  even  more  concerned about  our  health than you  are. We don’t  just eat one melon  a day,  rather:

(3)   We eat two  melons a day.

It is a fact  about  standard American or  British  English  that  we  cannot say:

(4)   *We eat two  melon  a day.

However, if we  were  speaking Indonesian or  Japanese, we  would say the  equivalent of two melon (three melon, four melon, etc.) because these languages don’t  use morphological plurals in sentences like this.

(5)   Indonesian:
Saya makan dua  buah  semangka (se) tiap   hari
I           eat       two  fruit   melon        every  day
‘I eat two  melons every  day.’

Japanese:
mainichi  futatsu-no meron-o            tabemasu every.day two-       gen melon-obj eat.imperf
‘I eat two  melons every  day.’

The  morphological grammar of English  tells  us  that  we  have  to  put an -s on melon whenever we are  talking about  more  than  one.  This fact of English  is so  transparent that  native  speakers don’t  notice  it. If we happen to be speakers of a language without obligatory plural marking, however, we will notice  it because we are going  to have  a lot of trouble with  it.
We  have   now   observed  something  about   English   morphology.  If a word is plural, it takes  the  suffix  -s. Living  creatures don’t  eat  only melons, however:
(6)   The evil giant  at the top of the beanstalk eats two melons, three  fish, and  four  children a day.

Everyone agrees  that fish is plural, but there  is no plural marker. Children is also plural, but it has a very unusual plural suffix, -ren, plus an internal change:  we say [TIld-] instead of [Tajld]. In other  words, it’s not  always the case that  we mark  plural words with  an s-like thing;  there  are other ways  in which  we  can  mark  plurals. Native speakers of English  know this,  and  they  do  not  need  to  think  about   it  before  making a  plural.
Consider the following:

(7)   Today  they  claim  that  they  will  fix the  clock tower  by Friday,  but yesterday they  claimed  that  it would take  at least  a month.

In  this  example, we  use  two  different forms   of  the  verb  claim. One is  present tense,  and  the  other  is  past.  Again,  this  is  not  true  for  all languages. If we  were  speaking Vietnamese, for example, we  wouldn’t make  any  distinction between claim and  claimed – we  wouldn’t mark the  verb  at all. If we were  speaking Chinese, we would not  distinguish between claim and  claimed in  a  sentence like  this,  because the  adverb zuótian ‘yesterday ’ is sufficient to indicate past  tense:

(8)   jintian  tamen shuo tamen xingqi  wä ké yà xiu hÜo  zhonglóu, today  they say    they     Friday can    fix  well clock.tower ká shì zuótian     tamen què       shuo zhì shÜo xu yào yíge yuè
but       yesterday they            however say    at least    need             a          month
‘Today  they  claim  that  they  will fix the  clock tower  by Friday,  but yesterday they  claimed that  it would take  at least  a month.’

If we  were  to leave  out  zuótian  ‘yesterday’, we  would need  to use  the particle le after  the  verb  to show  that  the  action  took  place  in the  past. In  other  words, whether or  not  a speaker must  indicate past  tense  in Chinese depends on context.
Notice  what  happens in  English  when we  use  some  other  verbs  be- sides  claim:

(9)   Today  they  say . . . but  yesterday they  said . . . tell us          told us know             knew

That  these  verbs  and  others  do  not  add  -t, -d, or -@d to make  their  past tense  is an  elementary fact about  English  morphology. We’ll talk  more about  verbs  like these  later  in the chapter. The next  observation about  English  morphology has  to do  with  pro- nouns. The  following is a real  exchange between an  American mother and  her  6-year-old son:

(10)   Who just threw a pool  ball through the basement window? Not  me.
In this context,  the 6-year-old (who was indeed guilty)  would never  have responded Not I. But if he were  to answer with  a sentence, the response would be  I didn’t,  not  Me  didn’t.  In  that  case,  the  object  form  of  the pronoun would be ungrammatical. Without formally knowing anything at  all  about   subjects   and   objects,   English-speaking 6-year-olds (and children even  younger) master the  pronoun system of their  language.
Given  the following sentence, how  many  children does  Joan have?

(11)   All  of Joan’s  children are  brilliant and  play  musical instruments surpassingly well.
From  this statement you  cannot  know  how  many  children Joan has, but one  thing  is certain:  she  has  more  than  two.  If Joan  had  only  two  chil- dren,  we  would normally say  both of Joan’s children, because it is a fact about  English  that  there  is a morphological distinction among universal quantifiers between the  one  designating all of two  (both) or all of more than  two  (all ) of a particular type  of entity.  In  some  other  languages, marking for  dual  is  even  more  pervasive. This  is  the  case  in  Ancient Greek,  as shown by the following examples:

(12)   ho stratiô:tes     lambánei tous             híppous the.nom.sg  soldier.nom.sg take.3sg   the.acc.pl horses.acc.pl
‘The soldier  takes  the horses.’
to:        stratió:ta         lambáneton tous         híppous the.nom.du  soldier.nom.du take.du            the.acc.pl horses.acc.pl
‘The two  soldiers take  the horses.’

hoi       stratiô:tai lambánousi tous     híppous the.m.pl soldier.pl take.3pl the.acc.pl horses.acc.pl
‘The soldiers (three  or more)  take  the horses.’
While  English  does  not  have  special  affixes  to  mark  the  dual,  it keeps track of the distinction through words like all and  both. There are actually languages in the world like Manam (Papua New Guinea:  Gregersen 1976) and  Larike  (Central Maluku, Indonesia: Laidig  and  Laidig  1990) that distinguish not only  singular, dual,  and  plural, but  trial  as well. The use of singular, dual,  trial, and plural second person subject prefixes in Larike is illustrated below:
(13)   Ai-         rala     iter-     lawa    pe?a-   o   ?2sg.sub- chop.down 1pl.incl.sub- garden finish-  qm
‘Did you  (sg.) finish  clearing our  garden?’
Kalu au- ?anu,  irua musti     iruai-   ?anu  si?u. if 1sg.sub- eat     2du certainly 2du.sub- eat    also
‘If I eat, certainly you  both  will eat too.’
Kalu iridu-ta- ?eu, au- na-   wela.if 2tri.sub- neg- go1sg.sub-irr- go.home
‘If you  three  don’t  want  to go, I’m going  home.’
Memang iri-   hise  tapi  imi- ta-       ?ari?i- truly    3pl.nonhum-  exist but   2pl.sub-neg-  see-ri.3pl.nonhum.obj
‘They really  do exist, but  you  (plural) didn’t see them.’

1.4       Background and Beliefs
In this text  is a general introduction to  morphology and  morphological analysis  from   the  point   of  view   of  a  morphologist.  The  purpose  is not to advocate any particular theory or to give the truth (whatever that is), but  rather to  get  you,  the  reader, to  where you  can  look  for  it by yourself. Still, it is inevitable that  some  of our  remarks will  be colored by  our  own  beliefs  and  background. We  would therefore like  to  pre- sent  some  of  our  foundational beliefs  about   linguistics and  linguistic methodology.
First,  we  believe  that  languages differ  from one  another.  You might be thinking, “Of course  they do!” But we mean  this in a very special way. Some  linguists are  always looking for ways  that  languages are  similar, and  at times, we do that, too. But we believe that if you focus only on the similarities between languages, you  miss  out  on all of the exciting  ways in which  they differ. What’s more, you may find parallels and similarities where none  really  exist.  We try  to approach linguistic analysis with  as open  a mind as possible, and  to do this, it is first necessary to appreciate the uniqueness and  diversity of the world’s languages.
Our  second foundational belief is that  languages,  which  we can write with  a small  l, are different from Language, with  a capital L. There  are thousands of individual languages in the world. But we may  also speak of language in  general to  mean  the  general phenomenon of Language that   encompasses all  individual  languages. This  Language is  related to  Noam Chomsky’s notion of Universal Grammar, which  posits  that languages are  all alike  in basic  ways.  There  is an  important distinction between these  two uses of the word language and  each is equally import- ant to linguistics. Individual languages have  features that are not charac- teristic  of Language in  general. For  example, one  feature of English  is that  its regular way  of forming plural nouns is to add  /z/. We would never  claim, however, that  this is universally true,  or that  it is a property of Language. To tie  this  belief  in  with  the  preceding one,  we  strongly believe  that  morphological theory and  morphological analysis must  be grounded in morphological description. If we  want  to appreciate what morphology really is, it’s best to have some idea of what  the morphology of individual languages is like. At the same time, we must  have a reason- ably well-thought-out general theory of the morphology of Language, so that  we can compare our  descriptions of individual languages within a wider context.  In  short,  linguists need  to  pay  equal  attention to  both small-l  language and  capital-L  Language.
Our  next  belief  is that  morphology is  a distinct  component of  lan- guages  or grammars. If you  are  not  already familiar with  some  of the controversy surrounding  morphology, this  needs   an  explanation.  The fact  that  some  languages, such  as  Vietnamese, do  not  have  morpholo- gically complex words has led some people to conclude that morphology should not be a separate branch of linguistics. The reasoning is that  lin- guistics is generally understood to deal with  properties of all languages – more  precisely, Language with  a capital  L. If there  are  languages that don’t  have  morphology, then  morphology is not  a property of all lan- guages, and  morphological phenomena should be treated in syntax  or phonology. We disagree. It has been shown elsewhere (e.g., Aronoff  1994) that  there  are aspects  of morphology that  cannot  be attributed to syntax or phonology, or anything else.
One piece of evidence that morphology is separate from syntax,  phono- logy,  and  other  branches of linguistics is that  words in some  languages are  grouped into  largely  arbitrary classes  that  determine their  forms  in different environments. Latin  nouns fall into  five  distinct classes,  called declensions, which  have little or nothing to do with syntax  or phonology, and  certainly cannot  be explained by  either.  They  are  purely morpho- logical  in their  significance. The uniquely morphological nature of these classes  is truly  brought home  by the  fact that  Latin  nouns also  fall into syntactic agreement classes (usually called genders) and  the two systems cross-cut one another: two  nouns may  belong  to the same  gender but  to different declensions and  vice  versa.  We’ll examine cases  like  these  in later  chapters, but  their  mere  existence  in  many  languages shows  that morphology must  be  given  some  independent status in  linguistics. Of course,  morphology, probably more  than  any  other  component of lan- guage,  interacts with  all the rest, but  it still has properties of its own.
We  also  believe  that  morphologies are systems. This  is  a  very  old observation. Because  of it, it is impossible to talk  about  isolated facts in a  language – everything holds   together. This  belief  together with  the second one, above,  are the reasons why  we’ll be looking carefully at the morphology of a particular language, Kujamaat Jóola,  throughout this book.  Considering the  morphology of  Kujamaat Jóola  in  close  to  its entirety will give us a valuable perspective that  we would never  gain  if we only  focused on isolated facts from  several  languages.
So far, we have given  you our beliefs about  the nature of language and morphology. We also have some that pertain to methodology. The first is that we should take an attitude of skeptical realism. Albert  Einstein  said that  a physicist must  be both  a realist  and  a nominalist, a realist  in the sense  that  you  must  believe  that  what  you  ultimately find  will  be real, but  a nominalist in  the  sense  that  you  must  never  believe  that  you’ve found what  you’re  looking for.  Martin  Joos  made a  similar  statement about  linguistics. On the one hand, you should always believe  that  what you are looking for is God’s truth, but on the other,  you should consider all that  you have  found so far as hocus-pocus. We believe  strongly in the value  of having a linguistic theory, but  we believe  equally strongly that you  should never  trust  it completely.
Our  other  methodological belief can be summed up  as a motto:  Any- thing  goes.  This  methodological belief  is  associated with  the  Against Method of Paul  Feyerabend, a twentieth-century philosopher who  felt that  if we  insisted on  a single  rule  of scientific  methodology, one  that would not  inhibit   progress, it  would be  “Anything goes.”  We  take  a no-holds-barred approach to  linguistics. We’ll use  any  tool  or  method that  will  tell us  how  language works.  This  attitude stems  in part  from our skepticism about  particular theories. People  who are wedded to indi- vidual theories tend   to  believe  in  using   tools  that  are  rooted in  that theory. Our  tools are not theory-based in that  way.  If a tool does  the job, we  are  happy to  use  it, whether it is a traditional linguistic tool  (e.g., native  speaker consultants, dictionaries, written grammars), an  experi- mental tool (e.g., imaging technology), or a statistical tool.
source: morphology and morphological analysis,page1-11,Mark aronoff and Kirsten fudman