Topic:Linguistics assignment
Subject:Linguistics
Volume: 2 pages
Type: other
Format: MLA
Description
Introduction to Linguistic Typology
Chichewa (Niger-Congo (Bantoid): Malawi)
(29) mlenje mmôdzi anabwérá ndí míkôndo
m-lenje m-môdzi a-na-bwérá ndí mí-kôndo
i-hunter i.sm-one i.sm-past-come with iv-spears
‘One hunter came with spears.’ (Mchombo 1998: 518)
In (29) the grammatical markers for noun class (i m/a and iv mi) and past tense (na)
are bound and are relatively straightforward to segment into morphemes.
Concatenation is a very frequent process indeed: a full 125 (75.8%) languages in
Bickel & Nichols’ database make use of concatenation exclusively for case and TMA
marking. As mentioned above, some languages combine concatenation with isolation.
Seven languages combine concatenation with non-linearity: two (1.2% of the sample)
Nilo-Saharan languages (Maasai and Nandi in Kenya, both Nilotic) combine concatenation
with tone while five (3%) languages, Hebrew, Egyptian Arabic (Semitic: Egypt),
Middle Atlas Berber (Berber: Morocco), Beja (Beja: Sudan) (all Afro-Asiatic) and
Lugbara (Nilo-Saharan (Moru-Ma’di): Uganda), combine concatenation with ablaut.
5.2.1.3 Non-linear markers
Non-linear markers involve some kind of modification to the host stem and are, as
the term implies, not straightforward to segment into chains of morphemes. There are
a number of ways that languages modify their stems nonlinearly. A very well-known
strategy is found in Semitic languages, where a root consists only of a set of consonants
(usually three) and where grammatical information is conveyed through insertion of
a pattern of vowels, commonly termed the “root-and-pattern” (Ussishkin 2006: 37)
but which is termed ablaut in Bickel & Nichols (2011b). Neither the root nor the vowel
pattern can function on its own. Modern Hebrew is a language with such a pattern;
an example of one kind of conjugation is the group belonging to the so-called pa’al
verbs, as shown in (30).
Hebrew, Modern (Afro-Asiatic (Semitic): Israel)
(30) g-d-r ‘enclose’
past: a-a (CaCaC): gadar ‘enclosed’
present: o-e (CoCeC): goder ‘encloses’
future: yi-Ø-o (yiCCoC): yigdor ‘will enclose’
imperative: Ø-o (CCoC): gdor ‘enclose!’
infinitive: li-Ø-o (liCCoC): ligdor ‘to enclose’ (Glinert 1989: 471)
In the above example, the root consonants remain the same, but the stem is modified
for tense through a set of vowel combinations, none of which can be segmented into
a linear string of morphemes. To indicate past tense, the root has to be modified with
the vowels a-a to form the stem gadarâ•‚; to indicate present tense, the root has to be
Chapter 5. Morphology 99
modified with the vowels o-e to for the stem goderâ•‚; and so on. This grammatical information
is not easily segmentable into a string of affixes, which makes it a non-linear
process. Agreement affixes (for instance -ti for ‘1sg’) may then be added concatenatively
to the various stems.
Another example of ablaut (also called gradation or vowel gradation) is found in
the strong verbs in Germanic languages, where inflection is marked through changes
in the root vowel quality, as in English sing – sang – sung (present – past – past participle).
Again we are not able to readily segment the words into a string of morphemes
along the lines of sing-past or sing-past.ptcpl since the grammatical information is
given through modifying the stem.
Ablaut is quite rare cross-linguistically; none of the languages in Bickel & Nichols’
sample make exclusive use of ablaut, while only five combine ablaut with concatenation
(and none combines ablaut with isolation).
Suprasegmentals (or prosodic formatives in Bickel & Nichols 2007), involving
tone, stress and length, are another type of non-linear morphological processes. Tone
is a well-known morphological strategy, common in continental South East Asia and
in sub-Saharan Africa. An example of a language with grammatical tone is Lango.
Lango (Nilo-Saharan (Nilotic): Uganda)
(31) a. àpônnê
1sg.hide.pfv.mid
‘I hide myself.’
b. ápònnê
1sg.hide.prog.mid
‘I am hiding myself.’ (Noonan 1992:101)
The difference between the perfective in (31a) and the progressive in (31b) is indicated
through change of tone: a falling tone on àâ•‚ plus a rising-falling tone on â•‚ôâ•‚ for the
perfective versus a rising tone áâ•‚ plus a falling tone â•‚òâ•‚ for progressive. In neither case
are the words possible to segment into a linear sequence of morphemes.
While tone is a frequent feature in the languages of the world (see 4.2.3.2), not
many make use of tone to convey the grammatical information sampled in Bickel &
Nichols. Six languages (3.6%) make use of tone, three of them exclusively – Iau (Lakes
Plain (Lakes Plain): Indonesia), Kisi (Niger-Congo (Southern Atlantic): Guinea) and
Lango – one, as mentioned above, combines tone with isolation and two, also mentioned
above, combine tone with concatenation. It seems reasonable to assume that
the rarity of tone in the sample is partly due to the non-proportionality to language
density pointed out by Maddieson (2011f) and quoted above (4.2.3.2).
The use of stress to convey grammatical information can, for instance, be found in
Italian, as in the example mentioned above (4.2.3.3): /ˈcantɔ/ ‘I sing’ and /canˈtɔ / ‘I sang’.
100 Introduction to Linguistic Typology
An example of length serving as a morphological process can be found in the Agar Dinka
(Nilo-Saharan (Nilotic): Sudan) case system, where the difference between tôoc̰ ‘swampy.
area.absolutive’ and tôooc ̰ ‘swampy.area.locative’ (Andersen 2002:13) is only that the
locative form lengthens the vowel. In other words, there are no affixes to segment.
Replacement or substitution is when a regular marker replaces a part of the stem,
something which is common in Nilotic languages (Bickel & Nichols 2007:182). For
instance, in Lango â•‚ê is a common plural marker that attaches to the stem. However,
if the stem ends in a vowel, the final stem vowel is replaced by the plural marker.
Compare, for instance Example (32), where â•‚â in bʊ́râ is replaced by the plural â•‚ê.
Lango (Nilo-Saharan (Nilotic): Uganda)
(32) singular plural gloss
rɛ́c récê ‘fish’
bʊ́râ bùrê ‘cat’ (Noonan 1992: 83)
Another type of replacement is suppletion, where a root or stem is paradigmatically
replaced by a root or stem of a different etymological origin. For instance, in English
the verb to go is inflected for past with a completely different stem, went, which is not
a cognate (does not have the same historical origin) with go.65
(33) She goes to school. (present tense) ~ She went to school. (past tense)
A rare type of non-linear process is subtraction, where the grammatical information
lies in taking out an element of the stem. For instance, in Nuer (Nilo-Saharan (Nilotic):
Sudan) some plurals are formed by shortening the vowel of the stem, as in ka̱a̱t ‘vulture.sg’
versus ka̱t ‘vulture.pl’ (Wright 1999: 33). In O’odham (Uto-Aztecan (Tepiman):
USA) the perfective form is regularly derived by subtracting the final consonant from
the imperfective form, as in hi:nk (imperfective) versus hi:n (perfective) ‘bark’ (Bickel
& Nichols 2007:183).
5.2.1.4 Reduplication
Reduplication falls somewhere in between concatenation and non-linear processes.
Because the languages many of us are used to, such as western European languages,
do not productively employ reduplication, we might not be aware of how common a
strategy it is. A full 84.7% (312 of 368 languages) of Rubino’s (2011) sample has productive
reduplication. English is part of the minority group of 56 languages which has no
productive reduplication. Map 5.1 shows the patterning of reduplicating versus nonreduplicating
languages in Rubino’s sample.
65. For a detailed discussion on suppletion, see Veselinova (2006).
Homework 4: Due Friday 3/9/2017
Part 1: Morphological Types
Problem 1: Morphological Analysis of Chitumbuka Verbs
Chitumbuka is a Bantu language of Southeastern Africa. It is a very synthetic language and though
it’s relatively agglutinative, it has some synthetic elements. Below is a verb paradigm for the verb
meaning “to beat” in the recent past.
“has
beaten”
Singular
Objects
Plural
Objects
1st Person 2nd Person 3rd Person 1st Person 2nd Person 3rd Person
Singular
Subj
1st Pers. ninitimba nikutimba nimutimba nititimba nimutimbani niŵatimba
2nd Pers. wanitimba wakutimba wamutimba watitimba wamutimbani waŵatimba
3rd Pers. wanitimba wakutimba wamutimba watitimba wamutimbani wakuŵatimba
Plural Subj
1st Pers. tanitimba takutimba tamutimba tatitimba tamutimbani takuŵatimba
2nd Pers. mwanitimba mwakutimba mwamutimba mwatitimba mwamutimbani mwaŵatimba
3rd Pers. ŵanitimba ŵakutimba ŵamutimba ŵatitimba ŵamutimbani ŵaŵatimba
Answer the following questions about the paradigm above:
1) What is the root?
2) List the subject-marking prefixes and provide the gloss you would give to each of these
morphemes in a three-line gloss.
3) List the object-marking prefixes and provide the gloss you would give to each of these
morphemes in a three-line gloss.
4) One of the object-marking morphemes is not like the others. Which morpheme is this and
why? What is the technical name for a morpheme of this part (Hint: check the end of the
word)
5) What aspects of this paradigm are agglutinative?
6) What aspects of this paradigm are fusional?
Problem 2: Morphological Analysis of Warlpiri Noun Declensions
Warlpiri is a Pama-Nyungan language spoken in Central Australia and is categorized towards the
fairly synthetic and highly agglutinative. This language inflects its nouns for case and number but
some of the “levels” of case and number you may be unfamiliar with. Warlpiri marks its nouns as
singular (exactly one), dual (exactly two), paucal (a small group) and plural (a large group). The
language also marks several cases, of which I’ve given you the six most common. These are ergative
(subject of a transitive verb), absolutive (subject of an intransitive verb or direct object of a
transitive verb), dative (usually an indirect object), allative (to, up to), elative (from, out of), and
locative (at, on, in). Below you will find the declension pattern for the Warlpiri word for “child”:
“child” Singular Dual Paucal Plural
Ergative kurdungku kurdujarrarlu kurdupaturlu kurdungku
Absolutive kurdu kurdujarra kurdupatu kurdu
Dative kurduku kurdujarraku kurdupatuku kurduku
Allative kurdukurra kurdujarrakurra kurdupatukurra kurdukurra
Elative kurdungurlu kurdujarrangurlu kurdupatungurlu kurdungurlu
Locative kurdungka kurdujarrarla kurdupaturla kurdungka
Answer the following questions about the declension pattern above:
1) What is the root?
2) Can you identify the suffixes meaning: singular, dual, paucal, and plural? (Hint: if there is no
suffix, assume the suffix is a “null” suffix, written: -∅)
3) Two of the number suffixes are pronounced in the same way. Which two are these?
4) If the two suffixes are pronounced in the same way, then why might we still argue that this
language considers these two separate number categories? No need to be perfectly
accurate, just list some things that might lead you to suspect that even though these two
categories are marked in the same way they are actually distinct. (Hint: if you’re stuck,
think about (a) semantics and (b) other types of words that might be marked for number
according to other patterns)
5) Can you identify the suffixes meaning: ergative, absolutive, dative, allative, elative, and
locative?
6) Two of these cases have somewhat unusual patterns. Which two cases showed this unusual
pattern? Describe this pattern.
7) Are the patterns you described in (6) an examples of morphological fusion in Warlpiri?
Why or why not? (Hint: there is a better answer here, but either way is debatable)
Part 2: Basic Categories
Problem 3: Universal Lexical Classes
What are the two lexical class distinctions that Whaley argues every language makes? In other
words, what are the four universal lexical class categories? (Hint: one of these is a sub-distinction
within the other). Give an example of a word or words in each of these categories.
Problem 4: Roots and Parts of Speech
For each list of words given below, do the following. (1) Identify the root shared between the
words in this list, (2) For each of the words in the list, identify the part of speech, (3) Next, try to
identify the part of speech of the root. If you can, explain how you came to that conclusion. If you
can’t, explain why this list presents a problem.
After you’ve done this for all the lists of words, give me a short summary about what your Week 6
readings have to say about whether or not roots have a part of speech in all languages.
List 1: bake, baker, bakery
List 2: systematize, unsystematic, systematic, system
List 3: befriend, friendship, unfriend, friendly
List 3: head (to the store etc.), behead, head (on the body), heady,
List 4: play (football), play (theater), play (with children), play (a tape)
List 5: hospital, hospitalize, hospitable, hospice
Problem 5: Semantic Roles
Identify the semantic role of the noun in bold for each of the following sentences:
1) [Pablo] chased [Boyo] around [the living room].
2) [Inky] is afraid of everything.
3) [Jordan] fed [Hugo] [a bowl of milk]
4) [Rachel] sent [cookies] to [her brother]
5) [Sandra] fled [Chicago]
6) [Quentin] hates eating [papaya]
