As well as a professor of
Linguistics, Anatole Lyovin (right) is also a priest in the Orthodox Church
in Honolulu, Hawaii.
These projects were produced for the course in 2005:
Christopher Maden: Manchu
Kevin Ng and Stanley Chan: Dialect
of Chenghai (Chiu Chow)
Ho Chuen Hei, Leung Lok Yan & Mau Chun Hong: Catalan
Lok Wing Hang: Franco-Provencal
Wong Yiu Hang, Corwin: Languages
of the Eskimo
Jo-jo Cheung & Erica Ling: Thai
Rachel Li: Breton
Vanessa Lam: Catalan
Lau Chaak Ming and Sabena Lam: Ainu
Chan Chui Yan & Yeung Wing Tai, Philip: Swahili
Chan Ching Yi: Icelandic
Wong Ka Man, Kevin: Sanskrit
Silvia Leung: Pashto
Jackie Kam and Mikalin Lau: Swahili
Mak Wing Yan: Hindi:
(not accessible on www)
Eric Tang & Shirley Kung: Shanghainese
(not accessible on www)
Q: I would like to know what case relations are?
A: These are the grammatical relationships between elements of the
sentence like subject, object and possessor. They are called case relations
because in languages that have case (like German or Japanese) they are
indicated by case forms (nominative, accusative, genitive etc).
Q: What does it mean by the question "how does it signal case relations?"
A: It suggests that those relationships which are indicated by case
in some languages (Latin, Russian etc: see lecture
3) may be indicated in different ways in other languages.
Q: I would like to ask Question 3 which is about Huave: what does the
"general language type" mean?
A: This is an invitation to categorise the language in terms of typology.
In the book and the classes we have talked about:
(i) word order typology,
(ii) morphological typology.
Since you have probably dealt with word order type in (i) already,
it is probably asking about (ii).
Q: In today's lecture, Hawaiian is said to have several particles marking different tenses and aspects, etc. I am confused about what exactly particles are. In English, a particle is an adverb or a preposition that can combine with a verb to make a phrasal verb, as "up" in "She tore up the letter". What about in other languages? Also only for constituting phrasal verbs too?
A: It happens that this class of words are called particles in English, where phrasal verbs are also known as 'verb-particle constructions'. But in the descriptive traditions for other languages the term is used for many other things, while things like verb-particle constructions may be treated differently. The German counterparts are called separable prefixes, as in auf-bauen "to build up", while hei as in lo hei 'pick up' or dai as baai dai 'put down' are called complements (buyu) in the Chinese tradition. Trask (1993) gives an unusually apologetic definition for 'particle':
"A label typically applied to some more-or-less well defined class of uninflected words in the grammar of some particular language when no more obvious label presents itself".
Similarly, Crystal (1999) defines particle as: "An invariable
item with a grammatical function. The term is especially used for a form
which does not readily fit into a standard classification of parts of speech".
On Cantonese (Oct 18, 2005)
Q: After today's lecture, I have a couple of questions in my mind. First,
I can't follow why gat-zat "cockroach" is a trace of prefixes.
You mean gat- is attached to the stem zat, which can stand alone to mean
something?
A: 'trace' means that it's a relic of what used to be a prefix.
The hypothesis is that it used to mean something, but if it was a prefix
it would not have stood alone. The evidence is that a prefix /ka-/ is found
in insect terms in other dialects, e.g. ka-lauh 'flea' in
Xiamen (L.Sagart, Vestiges of Archaic Chinese Derivational Affixes in Modern
Chinese Dialects. In Sinitic grammar: synchronic and diachronic perspectives,
edited by Hilary Chappell. Oxford University Press, 2001).
Q: Another question is that according to the handout, Sinitic languages,
including Yue, have at most 8 tones. But from a Chinese course I took last
year, there are 9 tones in Cantonese, aren't they?
A: To say that a language has x tones always involves choosing a particular
analysis for variant tones. For example, in Cantonese the high level (55)
and high falling (53) tones as in san1 "new" are variants,
but do not contrast systematically so we treat them as allotones of tone
1. We then have 6 (or 9) tones, rather than 7 (or 10).
The 9-tone analysis comes from the Chinese tradition which treats 'entering' tones as separate, but linguists typically regard them as allotones of the high, mid and low level tones when these happen to occur in 'checked' CVC syllables. There is good phonetic evidence for this analysis: the tone contours of the three 'entering' tones are exactly what you would expect if the speaker were beginning to produce the corresponding level tone but it is cut short. For example, jip3 "to pickle, marinate" has a shortened version of the same pitch contour as ji3 "idea". See p. 21 and p. 29 of Bauer's article 'Hong Kong Cantonese tone contours' in Studies in Cantonese Linguistics, edited by S.Matthews (Linguistic Society of Hong Kong, 1998).
Note that the traditional Chinese analysis might be justified for some dialects, it's just unnecessary for Cantonese. The big mistake that Chinese scholars make, in my view, is to assume that the same system somehow applies to all kinds of 'Chinese' even though they are mutually untelligible languages. These languages are at least as different as French, Italian and Spanish, related languages where nobody would suggest that they somehow follow the same phonological or grammatical system. This question is discussed by Matthews and Yip in Sinitic grammar: synchronic and diachronic perspectives, edited by Hilary Chappell.
Q1: According to your lecture 3 note, vowel harmony is something
about the front vowels versus back vowels of a word. I wonder if this is
the only rule that determine the vowel harmony.
A: No, there can be other rules. Please see the discussion on "Vowel
features and Vowel Harmony" below.
Q2: For agglutinating language like Korean in which word is made up
of certain morphemes, to consider whether the language have vowel
harmony, should we consider the whole set of morphemes (say, mikuk-ese)
or just focus on each separate morpheme?
A: You need to look at the whole word because if VH applies it generally
applies throughout a (complex) word. But if you see the same morpheme appearing
with different vowels (e.g. sometimes as -asa and sometimes
as -ese) that would suggest VH too.
Q3: Words of some languages may contain certain indicators/sounds that
we have no knowledge about the pronunciation of them, so how should we
infer the rules of vowel harmony of the languages in this case?
A: As in assignment 1, you can assume that the letters have the IPA
values unless you are told otherwise. The book tells you quite a lot about
the sounds of the particular languages involved, e.g. Turkish vowels on
p. 104, Tibetan sounds on p.151.
Assignment 2: Terms and abbreviations (Oct 12, 2005)
Q: Would you explain what "definite object", "characterising suffix", and "topic marker" are?
Q: I would like to ask which part(s) of the chapters of the textbooks
can be skipped. When I read the assigned reading chapter, I discovered
that some chapters involve certain topics that are not discussed or just
briefly mentioned (such as language families) in the class. Are we obliged
to read all the parts of the chapters. Or if possible, can you give us
a list about the parts/ chapters that can be skipped.
A: Please follow the syllabus
page where the relevant chapters are shown in parentheses. You can
skip chapter 2 on writing systems. I'm not sure what you mean about language
familes. When Lyovin focuses on a particular language such as Finnish or
Tibetan (his "sketches"), you do not need to follow every detail.
But note that:
(a) Tibetan features in the second assignment, so the sketch of Tibetan
my be useful there;
(b) the sketches make good models to follow in your language projects.
As well as a professor of
Linguistics, Anatole Lyovin (right) is also a priest in the Orthodox Church
in Honolulu, Hawaii.
Attendance and participation (Oct 5, 2005)
Q: It is stated in the course website that 20% of the final grade is
on attendance and participation. What is the ratio of one to another? Say
10% for attendance at the 4 tutorials and 10% for the participation?
A: The 20% is not divided, but attendance means marks may be deducted
for missing tutorials without giving a good reason, for example.
Q: And what accounts for the participation part? I wonder if it simply
refers to the tutorial performance or it also includes discussion during
lectures as well as by e-mail and on the discussion page.
A: Yes, all these count as participation: for example, some people
who may not talk a lot in tutorials may prefer to send questions by e-mail,
or ask them outside/after class. This is stated on the home page for Semantics,
but so far not on the Languages of the World page, so thanks for raising
this.
Q: In the lecture this morning, you mentioned some implicational universals,
which are normally associated with SOV languages. One of those is that
SOV languages tend to have postpositions. I looked the word 'postposition'
up from the dictionary and I am confused about the given definition: "a
word or part of a word that comes after the word it relates to, as -ish
in greenish".
A: This is a non-technical definition and not accurate: -ish
is a derivational suffix deriving adjectives.You need to look up technical
terms in a dictionary of linguistics such as:
A Dictionary of Grammatical Terms in Linguistics, by Larry
Trask
or: The Penguin Dictionary of Language, by David Crystal.
Crystal defines postposition as "A word the follows a noun phrase (often a single a noun or pronoun) to form a structural constituent, often of adverbial function. The notion is analogous to the use of a preposition in front of a noun phrase".
Q: I am not sure if the use of postpositions and the use of suffixes
in agglutinating languages we came across last week are actually the same
thing.
A: Basically, the difference is that a postposition is a separate word
(but forms a phrase, PP, with the preceding NP) while a suffix is part
of a preceding word. The difference is clear in Hungarian where the ablative
meaning ("away from") is expressed by a case suffix (-tol or
-töl, where the vowel harmony shows it is a suffix, see lecture
2a) but the meaning "beside" is expressed by a postposition, mellett:
senki-töl
a ház mellett
nobody-from the house beside
"from nobody" "beside the house"
In some languages it is more difficult to tell, e.g. in Japanese, ni
"to" behaves sometimes like a case marker and sometimes like a postposition.
In linguists' example sentences, you can "tell" a postposition by the fact
that it is written as separate word, whereas case suffixes are written
continuously as part of the word or separated by a hyphen, as in Lyovin's
examples.
Alternative terms (Sept 30, 2005)
Q: All the way we have stressed that under genetic classification, languages are grouped into families proven by sound correspondence in basic vocabulary. But I wonder how we can do so in some special situations, say Hong Kong. The thing is, in Hong Kong we speak Cantonese but write in Chinese using Chinese characters. In other words, there is a situation where an object verbally named may be different from the form we write on paper. For example, "tongue" can be said verbally as lei but instead as sit when it is written down. So a linguist should take the sounds of lei or sit in an attempt to genetically categorize the language Cantonese?
A: As you know, synchronic linguistics focuses primarily on speech rather than writing. But in such cases there are two forms which can be spoken. If you have two words for the same meaning, either one could be related to other languages by sound correspondences. And since written language is typically more conservative than spoken language, the forms used in written language are often older, and hence more useful in historical comparative linguistics. We had an example in Lecture 1, where muk is a literary word for 'eye' in Cantonese, and this is the word which can be related to Tibetan mig, Burmese myak, etc.
Sound correspondences and basic vocabulary (Sept 26, 2005)
Q1) On p9 of the book, there is an example showing
that there are sound correspondence in nonbasic vocabularies. In doing
the ex, i really not sure whether it is a must to search for basic vocabularies
in Finnish and Hungarian when establishing genetic relationship.
A: Well spotted! There is nothing to say that
there cannot be sound correspondences in nonbasic vocabulary. It's a question
of methodology: if we don't know whether the languages are related or not,
we cannot trust the non-basic vocabulary, because such words are
often (though not necessarily) borrowed.
Q2) In the discussion page, you said sound correspondence
refer to the articulation of sound. However, i do not know how to pronounce
Finnish and Hungarian (i am a year two student and haven't studied the
core course teaching phonetics), how can i know the sound correspondences
in the 2 languages?
A: Good point. Assume that the letters are pronounced
similar to the IPA equivalents. The footnote at the bottom of page 25 tells
you about some exceptions.
Q3) In doing the exercise, I really do not know
how to decide whether a word is a loan word. For example, zero is a loan
word in English, but i do not know whether it is a loan word in Finnish
and Hungarian. Are there any methods to know if a word is a loan word?
A: If you know one language (English) where that
word is borrowed, then it could be borrowed in any other language.
You probably won't be able to find out whether it is actually a loan or
not, so if you use it as evidence, use it very carefully (e.g., say that
this item may be misleading as it is a possible loan word).
Q4) I want to ask if the page about midterm quiz
is an updated page? As the home page written that the quiz is weighed 30%,
but on the note you have given to us, the quiz only weigh 25%. Is the information
on that page about midterm quiz relevant and useful to us?
A: Sorry, the home page has not been updated.
This year's handout is correct: the 3 assignments make up 30% and the quiz
25%.
Q5) Finally, i want to ask how often should we
surf for the discussion page? I surfed for it last Tuesday and then i surf
for it today. There really a lot of information covered on the page and
i feel confused.
A: Normally once a week would be enough, but
most of the questions seem to come just before the assignments are due!
Note that you may need to 'reload' to ensure that you are looking at the
latest version.
Sound correspondences and basic vocabulary (Sept 26, 2005)
Q1) What does sound correspondence refer to? Does
it refer to the manner of articulation of particular letter in the words
from 2 languages or does it focus on the similar spelling and the pronunciation
of the whole word from two languages? Would you explain it with some examples?
A: Spelling is not directly relevant, because
the same sound can be written differently in different languages, and vice
versa. So we need to focus on the articulation of sounds. There are examples
in exercise 1 on p. 25: the sound [s] is written 's' in Finnish and 'sz'
in Hungarian. So these spellings represent the same sound, and could be
a sound correspondence if they are (a) recurrent and (b) found in the basic
vocabulary.
Q2) How could we know/decide some sounds of genetically
related languages are preserved and inherited from the same ancestor language?
A: There are several steps involved: eliminate
loan words, identify sound correspondences in the basic vocabulary, etc.
(see lecture
1).
Q3) How could we differentiate sound change and
regular sound correspondences? I find that the latter one may also be caused
by sound change, like the example of English t-> German z-.
A: No need to differentiate, as sound correspondences
may (or may not) involve sound change. For example, m <-> m and
t <-> z are both sound correspondences between English and German.
Q4) When we classify whether 2 languages are genetically
related, we have to focus on the basic vocabulary. Is it because the basic
vocabulary of 2 languages is usually inherited from the common protolanguage,
but not borrow from other languages?
A: Yes - if the words are borrowed (or: might
have been borrowed) they are not evidence for a genetic relationship.
Q5) Why does the basic vocabulary include lower
numbers e.g.1,2,3 etc, but the higher and abstract numbers e.g. ten, one
hundred, zero etc, are excluded?
A: The principle is the same as for basic vocabulary
in general (see discussion below): it is assumed that all cultures use
the concepts "1" and "2" but not all use "0" or "100". For example, if
you count in base 12 or base 20, there would be no need for a word meaning
"100".
Q: Since I'm new to this course, I have not attended
any lecture yet...
A: Okay, you can download the lecture notes from
the syllabus
page, but it would be better to copy a friend's handouts with their
notes. You should also read chapter 1 of the textbook.
Q: How much are we expected to write for the answer?
A: We said about 2 pages in class. Please do
not write an essay!
Q: Should it be printed out in any specific format?
(e.g double line spacing)
A: I don't mind about line spacing, though in
general I object to unnecessary use of paper (such as cover pages). For
a data-based question like this, it is helpful to put the relevant data
into tables.
On basic vocabulary
Q: I have gone through chapters 1 and 3 of the core textbook by Anatole Lyovin and I have a question concerning the term "basic vocabulary". Although you have suggested a few domains which are considered basic vocabulary in the first lecture's notes(say kinship terms, body parts, lower numbers and pronouns), next week, I started to question what it takes to be basic. For example, is "cracker" basic? It is in some sense, for most people around the world must perhaps know what it is. It may not on the other hand if you argue that some very poor people in the third-world countries have never even had a taste of it. The same dilemma also applies to terms like "under", "soap", "ice", "soft", "egg", etc.
A: Good question! The ideal basic vocabulary item
is one where no culture can do without the concept. You have noticed that
some cultural groups might not be familar with items such as crackers and
soap. This is sufficient to cast doubt on their "basic" status. "ice" and
"egg" are interesting cases because whether they are basic terms would
seem to depend on the geographical location of the speakers. We can assume
that all cultures know eggs because even if they don't keep chickens, they
should be familiar with other birds and animals laying eggs. For "ice"
we really need to consider climate: basically, any culture outside the
tropics will experience ice, so the item should be basic in temperate (non-tropical)
climates.
Q: I would like to know if there are any required
readings for each lesson of LING 2009. Actually, I have got the textbook,
but I don't know which chapter I should read for each lesson. Would you
give us some guidance?
A: For the moment, read chapters 1 and 3. You
can ignore chapter 2 because we won't talk about writing systems. For the
rest of the course, there is usually a chapter corresponding to the area
of the lecture. These are now shown on the syllabus
page.
Objectives of the course (Sept 15, 2005)
Q: The content so far seems very factual. Are
we supposed to learn and memorise so many facts about particular languages?
A: No, the more important points are the principles
of
classification: on what basis do we recognize families such as the Romance
family? And why do we consider Cantonese, to be a Sino-Tibetan language?
Similarly for typological classification: what does it mean to be an agglutinating
language (see lecture
3) or an 'SOV' language? (see lecture
4). The same applies to the Midterm
Quiz.
Q: How long should the term paper be?
A: Aim for 10-12 pages. Similarly for web-based projects.
Q: How will the papers be assessed?
A: There should be an element of research in them, though this usually
means library and Internet searching rather than collecting your own data.
In writing a term paper you are unlikely to discover something altogether
new. However if you can discover things that nobody in the class knew before,
or that the teacher does not know, this should be cause for both satisfaction
and credit. Some general principles are:
Q: When is the paper due?
A: Monday, December 19th (2005) as given on the syllabus
page. As this is quite a late date, two weeks beyond the end of term,
there will need to be penalties for late submission.
Q: I find that the Swahili loanword from Arabic, adaba
"good manners" in exercise 3 is spelt as adabu in exercise
1. Which spelling is correct? I can only find adabu in an
online Swahili-English dictionary.
A: Well spotted! In principle this could be either a typo (misprint
or copying error) or dialect variation. The word does not seem to appear
in my Swahili phrase book. But you now have two sources for adabu
versus one for adaba. Maybe you can add a linguistic argument
that 'adabu' is correct?
Assignment no. 2: Exercise 2
Q: i got a question concerning the tutorial
assignment. Q2 asks us to analyse the Xu language. But i don't really
understand the term "case relations", which is in (Q2i). i know what case
system is. But what is case relations?? Is it something about the agreement
between two cases (like object and subject) in a sentence?
A: You're on the right track: case relations
means the relationships between noun phrases and verbs, such as subject
and object. These are relationships which are marked by case in languages
that have cases. There must be some way to distinguish the subject and
the object of sentence: if a language does not have morphological case,
then it may use word order (as in Chinese) or agreement (as in Swahili).
Q: I am writing to ask when will we need to hand
in our 2nd tutorial assignment and whether we should do it in a formal
way, i.e. quotation, footnotes, etc.
A: Details of the 2nd assignment are on the tutorial
page. These are problem-solving rather than essay-writing exercises,
but in either case it's important to cite any sources that you use: make
sure that you quote when using an author's exact words, and specify the
sources in footnotes and/or a reference list.
Q: I got a question when i was going over the lecture note 3. I found
the concepts of portmanteau morphemes and syncretism quite confusing. What
are the differences between them? Could you explain them in more detail?
A: Syncretism which is when forms 'fall together': the same
form is used for two or more distinct functions. For example, the Latin
word
puellis can mean "for the girls" (dative case) or "by/with
the girls" (ablative case).
Portmanteau (French for "hold-clothing") is an old-fasioned
word for a kind of suitcase. So a portmanteau morpheme holds several things
(features) inside itself. In English the 3rd person /-s/ is considered
a portmanteau morpheme because it carries the feature [present tense] as
well as [3rd person].
Both of these morphological phenomena are commonly found in fusional
languages, like Latin and other Indo-European languages. Some traces of
them are found in English, even though much of its morphology has been
lost.
Q: Which type of morphological typologies does English belong to? e.g
He
goes to school -- Is 'go' the stem? And '-es' is a suffix? Can
we say '-es' a morpheme?
A: Yes, the word 'goes' contains a suffix /-s/ which in this case is
written '-es' but pronounced [z]. This is a morpheme -- a meaningful unit
which contains the information [+3rd person singular] and [+present tense].
Because it contains two different features (person and tense) it is technically
a portmanteau morpheme, and these are typically found in fusional languages.
Q: How about Japanese? e.g. watashi wa gatsugo e ikimasu
I school to go (I go to school). To say ' i went to school', 'ikimasu'
has to be changed into 'ikimashita'. Does it mean '-masu' and '-mashita'
are the suffixes?
A: -mas(u) and -(i)ta are both suffixes, indicating politeness and
past tense respectively. The vowels in brackets have to be inserted to
make the words fit the syllable structure of Japanese, which is (C)V(N).
Q: Could the above examples prove that English and Japnanese are the
fusional languages? (I still wonder if the suffixes like '-es' in English
or '-mashita' in Japanese a morpheme because these suffixes do show meanings.
If they are morphemes, English and Japanese would be agglutinating languages.
Am I right?)
A: Japanese is agglutinating because you add one suffix after the other,
and each carries a single feature. This contrasts with the English examples
where the suffix /-s/ carries both person and tense.
Midterm Quiz: for details see the Midterm Quiz page.
Q: I found that there are some part of Chapter 3 which have not been
taught in lecture so far, do I still need to study those parts which have
not taught in lectures? Or I only require to study those parts from the
textbook which are equivalent to what have been taught in lectures?
A: That's right, there are several sections we haven't talked about
in the lectures. For example, in chapter 3 we basically covered up to p.
56. The best way to revise is probably to go through the lecture notes
and look up those topics in the book.
Q: Moreover, I realised that there many detailed information e.g. under
Indo-European there are many sub-branches, for these sub-branches do I
need to know and memorize each of their own characteristics or it is necessary
to have remember what sub-branches is under the Indo-European language
will do?
A: It should not be necessary to memorize particular facts, but it's
good if you can name some of the constituent families of Indo-European,
for example.
metathesis is a Greek term, literally 'cross-placement' (compare synthesis, 'putting together'). It means that two sounds have somehow changed places, for example in some English dialects, 'to ask' becomes 'aks'. And some time in the history of English, Old English hros (compare German Ross) became horse. The puzzle about this phenomenon is that unlike sound changes, it does not seem to be regular, but occasional and unpredictable. Consequently, in comparing words across languages, metathesis is much less useful than regular sound change.
Q: Besides, does basic vocabulary only include "kinship terms, lower
numbers, body parts and function, pronouns", or are there other categories
involved?
A: Yes, in particular, words for natural phenomena (wind, water, fire
etc) are considered basic vocabulary items too. There is a 100-word "Swadesh
list" of basic vocabulary which is made up of these words and the kinds
that you mention.
Q: Joseph Greenberg classified the African language into 5 types, i.e. the Afro-Asiatic, Nilo-Saharan, Niger-kordofanian, Khoisan and Linguae Francae. In fact, what does he base on to make such classifications?
A: Greenberg himself identified 4 families. The fifth category mentioned in lecture 7, 'Linguae francae, pidgins and creoles' is a mixed bag which we separated since it's not clear whether such languages can be said to belong to a family in the usual sense (this will be discussed in Languages in Contact next semester).
Greenberg intended his classification to be a genetic, not a typological one. However in the article on Khoisan languages by Tony Traill, he begins by pointing out that it has not been demonstrated that this is a single family. The use of click consonants as phonemes is really a typological similarity rather than a genetic one. However the fact that clicks are unique to this area, and perhaps also the cultural identity of the 'Bushmen', has encouraged linguists to believe that they are a single family. Notice that it would have to be a very ancient family, and it may be the case that the common ancestor, if any, lies too far back to be reconstructed.
More generally, Greenberg's method for classifying languages genetically differs from the classical Comparative Method. While the Comparative Method seeks to set up sound correspondences, Greenberg looked initially for superficial similarities in basic vocabulary among a very wide range of languages - a method which has been called 'mass comparison'. Greenberg's classification was controversial at first because most scholars were focused on smaller families such as Semitic, Bantu etc., where the correspondences were clear. Gradually however his classification became generally accepted, with the exception of Khoisan as noted above.
Greenberg's method became controversial again when he applied the same method in Language in the Americas (1987), arguing that the hundreds of recognized families could be grouped into just three; and in Indo-European and its Closest Relatives (2000) where he argues that Indo-European belongs to a Eurasiatic phylum along with Uralic, Altaic and other families. Superficial similarities are unreliable (see the question below on Sound correspondences and genetic classification) and because he was dealing with so many languages Greenberg inevitably made mistakes. However some argue that he has turned out to be right before -- notably in the case of Africa -- so even with its potential for errors, his method can sometimes come to the right conclusions.
Joseph Greenberg died in 2001, having just completed the second volume
of his book on the Eurasiatic hypothesis. One of his students recalls that
one of his last regrets was that he never got around to classifiying the
languages of Southeast Asia.
Q: Are there any different between the words "synthetic" and "agglutinating"? Or they are just adjectives describing the same things?
A: Synthetic just means "put together"
i.e. words are put together out of more than one element. It doesn't specify
how the words are put together, so a language or construction described
as synthetic could be fusional, agglutinating, or even polysynthetic.
Agglutinating means the words are
put together in a particular way, such that the individual parts remain
visible. It may be helpful to think of the term as meaning "glued together",
or in more Cantonese terms, to think of "glutinous rice", in which the
grains of rice are stuck together but can be clearly distinguished from
each other.
Q: Could you explain a bit more about subjective
conjugation and objective conjugation.
A: Conjugation is a traditional
terms for a series of different forms of the same verb, each form having
a different grammatical function. The terms "subjective/objective conjugation"
are used specifically for Hungarian and other Uralic languages where the
verb takes different forms depending whether it has a subject (i), or an
object too (ii):
(i) ker-ek
ask-1sg
"I ask"
(ii) azt ker-em
that ask-1sg-Obj
"I ask (for) that"
Q: Are there several types of vowel harmony (vowel
rounding, vowel tenseness, and the front-back vowel harmony)?
A: Yes (see below), though we have only looked
at front-back harmony in class since it is the simplest and probably the
most common type.
Q: I would like to ask what is vowel rounding
and vowel tenseness and what's the differences between ATR and RTR?
A: These are all phonological features which
vowels may have. They are relevant to us because they may be involved in
rules of Vowel Harmony. So far we have looked at the simplest and commonest
cases of Vowel Harmony: front-back harmony, which typically means that
every word must contain front vowels or back vowels but not both. This
means that the suffixes have to change to match the vowel of the word to
which they attach, as in Hungarian:
e.g. Istvan-nak "to Steven"
a haz-ban "in the house"
Peter-nek
"to Peter" a szem-ben
"in the eye"
Rounded vowels are those produced with rounded lips, like [y] in French 'tu', Cantonese syu1 'book', etc. In some languages, whether a vowel is rounded affects the choice of vowel in the rules of vowel harmony (e.g. Finnish, as described on p. 80-81 of the textbook).
Tense vowels contrast withlax vowels, e.g. the vowel in 'tin' is tense and lax in 'teen'. In English there is no vowel harmony but in some languages VH would require that a word has either tense or lax vowels.
Similarly, ATR (Advanced Tongue Root)
and RTR (Retracted Tongue Root) are articulatory features
in which vowels can contrast, and vowel harmony in terms of ATR/RTR is
found in some West African languages such as Yoruba. The distinctive feature
is usually ATR, so that a vowel can be [+ATR] or [-ATR]. However, the typologist
Joseph Greenberg proposed that some systems have RTR as the distinctive
feature. One analysis of Cantonese says that, for example, the allophones
of /u/ as in fu 'rich' and fuk 'blessing' differ in ATR or
RTR. Since the root (body) of the tongue is further back in fuk
(due to the following velar [back] consonant -k) the difference can be
described using the feature Retracted Tongue Root.
Q: The textbook seems to focus on initial consonants (i.e. those at
the beginning of the word) in looking for correspondences. Why is this?
A: This seems to be where the most consistent sound correspondences
can be found. The main reason is that phonetic erosion affects
principally the ends of words and syllables, so that final consonants tend
to be dropped or weakened. For example, many Chinese dialects (like Mandarin)
have dropped final stops, while Cantonese and southern Min dialects have
weakened them to unreleased consonants. In the Austronesian case (p. 9),
however, the consonant correspondences seen in initial consonants can also
be found in medial position (in the middle of the word), e.g. Samoan mata
corresponds to Hawaiian maka 'eye', an example of the same t/k correspondence
which you see in toa/koa, etc; and Samoan 'aso corresponds
to Hawaiian 'aho, enother example of the s/h correspondence as seen
in sal/hala, etc.
Q: What is meant by 'chance resemblances' and 'superficial similarities'?
A: There are a finite number of consonants and vowels in any language,
so there is always the possibility that an individual word just happens
to have a similar combination of consonants and vowels as the same word
in some other language. Historical linguists have recently tried to develop
statistical tests to distinguish chance resemblances from signficant ones,
but so far these methods remain controversial.
'Superficial similarities' (also called 'lookalikes') are pairs or sets
of words that look similar on the surface but are not part of a systematic
correspondence. Larry Trask in The History of Basque gives a list
of such superficial resemblances between Basque and Hungarian. Since we
have looked at both these languages it is interesting to consider a sample:
| Basque | Hungarian | |
| Noun Plural Suffix | -ak | -ak/-ek |
| bean | baba | bab |
| whale | balea | ba'lna |
| small, little | txiki | kicsi (with metathesis) |
| father | aita | atya |
| pair, couple | pare | pa'r |
| corner | zoko | zug |
This table, which Larry intends as a joke (his textbook Language: The Basics is very entertaining!) illustrates a number of possible reasons for similarity:
Q: (Why) are there only 4 morphological types?
A: There is nothing magic about the number four. The four types recur
in unrelated language families and in different parts of the world (for
example, predominantly isolating languages are found in Southeast Asia,
Oceania and West Africa). At least one more type can be identified, but
it seems to be found only in the Semitic family (Arabic and related languages).
In this 'introflecting' type, grammatical information is conveyed by the
choice of vowels. Arabic has have triconsonantal roots consisting
of three consonants, e.g. k-t-b "read". Forms of the verb
have different combinations of vowels and affixes, e.g. katab-a
"wrote", kutib-a "was written". See the full paradigm on
p. 211 of the textbook.
Q: Why is Type IV termed "polysynthetic / incorporating" - is it another
case of different terms for the same thing?
A : The textbook treats them as a single type but strictly speaking
they are not. The two tend to go together, but there are polysynthetic
languages that don't allow incorporation and vice versa. This is consistent
with the general status of a morphological type as a cluster of properties
which tend to go together.
Q: The textbook contains a lot of unexplained terms, such as nominals
(p.
19)...
A: If you find an unfamiliar term, an ordinary dictionary is unlikely
to be useful. In the case of nominal, for example, the everyday
meaning is qute different from the technical one. The best dictionary to
use is the Dictionary of Grammatical Terms in Linguistics
by Larry Trask (yes, him again!) which is in the reference section of the
UL at R 415.014 T7. He defines nominal as 'pertaining to
nouns or projections of nouns' (such as Noun Phrases - you could look up
projection
in
the same book)