Malayalam



Assignment #4: Malayalam

Due Wednesday, 31 Oct. 2007

The data below illustrate some alternations in Malayalam. Your job is to write rules to account for the alternations, as usual, but also to use the tools of lexical phonology to account for why some of the rules don’t apply in certain morphological contexts. You will have to make a proposal about how many levels there are in the lexical component of Malayalam phonology, decide which morphological operations belong to which level, and decide which rules apply at which level(s) of the lexical phonology, or in the postlexical phonology.

Malayalam is a Dravidian language but has been heavily influenced by Sanskrit. I’ve marked words as D (Dravidian origin) or S (Sanskrit origin) whenever the source gave that information (sometimes that information is phonologically relevant, sometimes not)—in some cases D could mean that the word is treated by the language as Dravidian even if it actually came from Sanskrit, and vice versa. In the cases where the etymology is not given, you may assume that D and S words behave the same. You may assume that lexical entries can be marked [+D, –S] or [–D,+S], and that rules can be sensitive to those features if they apply at a stage in the derivation before that information has been erased.

Be sure to discuss…

• At what stage in the derivation can or must morphological bracketing and/or abstract features like [D] and [S] be erased? (Remember that once information is gone, it’s gone—it can’t come back later in the derivation.)

• Do the lexical rules show any NDEB? If so, can it be analyzed using Kiparsky’s idea of blocking by identity rules?

A lot of the rules look kind of similar, which normally is a danger sign. In this case, avoid having very similar rules when you can, but don’t worry if your solution still includes some.

Data are from Mohanan (bibliographic info available later if you want). Some consonant allophony is not indicated. As always, the order in which you present your solution doesn’t have to reflect the order in which the data are presented.

Please don’t be alarmed by the large number of words! They’re there to help you see the patterns repeated, not because every word requires its own analysis.

1. In order for some of the data to make sense, you will first need an analysis of the following schwa/zero alternations. Analyze the glide/zero alternations also, for contrast. I am writing spaces or hyphens on either side of the glides so as to avoid committing to any morphological affiliation, but the glides do not represent a distinct morpheme—they are phonologically predictable.

Note: Even though they don’t form a natural class, [m] and [n] are the only consonants that are allowed to be word-final (on the surface) in this language.

There is something going on in 17—and in 106 and 143—that will not make sense until later.

|word | |inflected | |phrase | |modifier-head | |

|εωι?ε |where | | | | | | |

|ωαα〉↔ |sword | | |ωαα〉 εωι?ε |Where is the sword? | | |

|μυυρτ↑τ↑ακυυ??ι |sharpened | | |ωαα〉↔ μυυρτ↑τ↑ακυυ??ι |sharpened the sword | | |

|ωα〉α |bangle | | |ωα〉α ϕ εωι?ε |Where is the bangle? | | |

|υραΝΝι |slept | | | | | | |

|μαα♣↔ |teacher | | |μαα♣ υραΝΝι |The teacher slept. | | |

|ωαν5ν5υ |came | | |μαα♣↔ ωαν5ν5υ |teacher came | | |

|δ5αασι |servant | | |δ5αασι ϕ υραΝΝι |The servant slept. | | |

|αανα |elephant | | | | | | |

|καα?↔ |forest | | | | |καα??-αανα |wild elephant |

| | | | | | | |(lit. forest-elephant) |

|κυ|ι |hole | | | | |κυ|ι-ϕ-αανα |insect sp. |

| | | | | | | |(lit. hole-elephant) |

|τ5οοκκ↔ |gun |τ5οοκκ-ιντε |gun-poss | | | | |

|?ικκυ |Tickoo |?ικκυ-ω-ιντε |Tickoo-poss | | | | |

2. Some differences between D and S stems.

|word | | |modifier-head compound | | |

|ρϑατ5ι |Rati (name) | | | | |

|ϕακ♣ι |yakshi (demoness) | | | | |

|β⎥ααρϑϕα |wife | | | | |

|ωαδ5⎥υ |bride | | | | |

|μανυ♣ϕαν |man | | | | |

|πυστ5ακαμ |book | | | | |

|πυ♣παμ |flower | | | | |

|μαρϑαμ |tree, wood | | | | |

|ϕακ♣αν |yaksha (demon) | | | | |

|κυ??ι |child |D |ρϑατ↑ι-κκυ??ι |X-D |Ratikutti (name); ignore the [τ5]/[τ↑] difference—probably a mistake |

|πααλα |pala tree |D |ϕακ♣ι-ππααλα |X-D |the pala tree in which yakshi lives |

|ωιι?↔ |house |D |β⎥ααρϑϕα-ωιι?↔ |X-D |wife’s house |

|κυρϑαΝΝ↔ |monkey |D |μανυ♣ϕα-κκυρϑαΝΝ↔ |X-D |human monkey/ape |

|τ↑ααϕαμ |paint |D |πυστ5ακα-τ↑τ↑ααϕαμ |X-D |paint for books |

|κυ?αμ |pitcher |D |πυ♣πα-κκυ?αμ |X-D |pitcher of/for flowers |

| | | |μαρϑα-κκυ?αμ |X-D |wooden pitcher |

|κυτιρϑα |horse |D |μαρϑα-κκυτ5ιρϑα |X-D |wooden horse |

|παα??↔ |song |D |ϕακ♣α-ππαα??↔ |X-D |yaksha’s song |

|δ5εεωι |goddess |S |ρϑατ5ιι-δ5εεωι |X-S |goddess Rati |

|πααδ5αμ |foot |S |ϕακ♣ιι-πααδ5αμ |X-S |yakshi’s foot |

|γρϑ↔ηαμ |house |S |β⎥ααρϑϕαα-γρϑ↔ηαμ |X-S |wife’s house |

| | | |ωαδ5⎥υυ-γρϑ↔ηαμ |X-S |bride’s house |

|κα⎟?Ηαμ |neck |S |μανυ♣ϕα-κα⎟?Ηαμ |X-S |human neck |

|τ↑αρϑιτ5αμ |story |S |πυστ5ακα-τ↑αρϑιτ5αμ |X-S |history of books |

|κυμβ⎥αμ |pitcher |S |πυ♣πα-κυμβ⎥αμ |X-S |pitcher of/for flowers |

| | | |μαρϑα-κυμβ⎥αμ |X-S |wooden pitcher |

|τ5ααρϑα |Tara (name) | | | | |

|κααν5τ5αν |husband |S |τ5ααρϑαα-κααν5τ5αν |X-S |Tara’s husband |

| | | |τ5ααρϑαα-κααν5τ5αν-μααρϑ↔ |X-S-sf |Tara’s husbands (-μααρϑ = plural) |

|τ5ιι |fire | | | | |

|κα??α |lump |D |τ5ιι-κκα??α |X-D |lump of fire |

|ν5αα〉αμ |flame |D |τ5ιι-ν5αα〉αμ |X-D |flame of fire |

|σωαρ⎟⎟αμ |gold | | | | |

|πα|αμ |fruit |D |σωαρ⎟⎟α-ππα|αμ |X-D |fruit of gold |

|μα|α |rain |D |σωαρ⎟⎟α-μα|α |X-D |rain of gold |

|καρϑι |charcoal | | | | |

|κυυ??ααν |dish |D |καρϑι-κκυυ??ααν |X-D |coal dish |

|λα⎢⎢υ |sweet |D |καρι-λα⎢⎢υ |X-D |charcoal sweet (don’t worry about the [r]/[rj]—maybe a typo in the source) |

|μυυτ5ραμ |urine | | | | |

|σϑαΝκα |suspicion |D |μυυτ5ρα-σϑαΝκα |X-D |desire to urinate |

|συυτ↑ι |needle |D |σωαρ⎟⎟α-συυτ↑ι |X-D |golden needle |

|πο?ι |powder | | | | |

|μαλαρϑ↔ |popped rice | |μαλαρ-ππο?ι |X-D |powdered popped rice |

|πακαλ↔ |day | | | | |

|κινααω↔ |dream |D |πακαλ-κκινααω↔ |X-D |day dream |

|αα〉↔ |man | | | | |

|κυυ??αμ |group |D |αα〉-κκυυ??αμ |X-D |crowd |

|ωιρακ↔ |firewood | | | | |

|κο〉〉ι |twig |D |ωιρακ↔-κο〉〉ι |X-D |log for firewood |

|κα?υκ↔ |mustard | | | | |

|παατ5ραμ |vessel |D |κα?υκ↔-παατ5ραμ |X-D |vessel for mustard |

|καατ5↔ |ear | | | | |

|κυτ5τ5↔ |piercing |D |καατ5↔-κυτ5τ5↔ |X-D |piercing of the ear |

| | |D |καα??↔-κυρϑαΝΝ-αν |X-D |wild monkey; cf. 16, 39 (-an derives masc. sg. Ns) |

|τ↑αΝΝαλα |chain |D |μαρϑα-τ↑τ↑αΝΝαλα |X-D |wooden chain; cf. 31 |

3. Malayalam also has multiply headed compounds that have a coordinative meaning. Carefully compare the coordinative compounds below to the modifier-head compounds below and those already seen above; some of the differences are subtle.

|ωορδ | | |coordinative compound | | |

|υλαχχα |pounding stick | | | | | | |

4. Compare the data above to forms with inflectional and derived affixes:

|word | |inflected | |derived | |

|αωα〉↔ |she |αωα〉↔-κκ↔ |to her | | |

|αωαρϑ↔ |they |αωαρϑ↔-κκ↔ |to them | | |

| | |ακαλ-υμ |will recede |ακαλ-τ↑τ↑α |distance |

| | |ακαλ-υν5ν5υ |recedes | | |

| | |πακαρϑ-υμ |will spread |πακαρ-τ↑τ↑α |spreading |

| | |πακαρϑ-υν5ν5υ |spreads | | |

| | |κυ〉ιρϑ-υμ |will feel cold, cf. 136 |κυ〉ιρ-μμα |coldness |

| | |κυ〉ιρϑ-υν5ν5υ |feels | | |

5. Here is an alternation that applies to sonorants:

|word | | |inflected | |derived |

|μαα?↔ |cow | | | | |

|τ5ο|υτ5τ5↔ |stable | | | | |

|αα?↔ |sheep |αα?↔-μαα?↔-κα〉↔ |sheep and cows |αα??↔-τ5ο|υτ5τ5↔ |sheep shed |

|ωακκ↔ |edge | | | | |

|ροο?↔ |road | | |ροο??↔-ωακκ↔ |roadside |

|τ5ο|ιλαα〉ι |laborer | | | | |

|τ↑υμα?↔ |burden | | |τ↑υμα??↔-τ5ο|ιλαα〉ι |porter |

|κα??ιλ↔ |cot | | | | |

|καμπι |metal rod | | | | |

|καϕαρ↔ |rope |καϕαρ↔-καμπι-τ↑αΝΝαλα-ωαργγαμ |the set of rope, metal rod and chain |καϕαττ↔-κα??ιλ↔ |cot made with rope |

|κο??α |basket | | | | |

|τ↑αωαρ↔ |trash | | |τ↑αωαττ↔-κο??α |trash can |

|τ↑οορ↔ |rice | | |τ↑οοττ↔-παατ5ραμ |lunch box, cf. 101 (X+D!! make sure |

| | | | | |this form works) |

7. Some more phrasal data. Check that your analysis generates them correctly.

|word | |phrase | |

|μανυ♣ϕαν |man | | |

|μαρϑιτ↑τ↑υ |died |μανυ♣ϕαν μαρϑιτ↑τ↑υ |The man died. |

|κα?αμ |debt | | |

|κο⎟?↔ |with | | |

|μυ?ινϑνϑυ |destroyed |κα?αμ κο⎟?↔ μυ?ινϑνϑυ |We are destroyed with debt. |

|γαμιτ↑τ↑υ |went |τ5ααρϑα γαμιτ↑τ↑υ |Tara went |

|κααρϑα⎟αμ |reason |β⎥ααρϑϕα κααρϑα⎟αμ |because of the wife |

|ν5α?ι |actress | | |

|πρααρτ5τ5Ηιτ↑τ↑υ |prayed |ν5α?ι πρααρτ5τ5Ηιτ↑τ↑υ |actress prayed |

-----------------------

[1] You have limited data on what happens to nasal-obstruent clusters where the nasal doesn't delete (#123, 126, 148, 149, 158, 163, 228, 232, maybe a couple of others). So feel free to formulate fairly arbitrary-seeming rules (Νk does this, mk does that...), and focus on capturing the differences between different types of morpheme boundary.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download