Consider the following data from the Australian language Lardil. This language exhibits a process
known as truncation, a term which refers to the deletion of underlying material resulting in the
shortening of a form. Additionally, the language also exhibits a processes of augmentation, which
occurs when material is added to the underlying form for phonological reasons.
Please focus on the alternations which are most clearly visible between the underlying
stems and their corresponding nominative forms. Data from other categories (nonfuture
accusative, future accusative) is also provided, but your analysis needs to account for only the
nominative forms. “C” stands for “consonant” and “V” stands for “vowel”. Pay careful attention
to the way in which the data have been organized; to help guide you through the data, I’ve included
a brief summary of each set of data following its presentation:
a. C-loss from stem
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘story’
// [] [-] [-]
‘queen fish’
// [] [-] [-]
‘boomerang’
Here in (a), we see that in the nominative form of nouns, a consonant is lost from the end of the
underlying stem. Now look at (b):
b. V-loss from stem
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘oyster species’
// [] [-] [-]
‘rainbow’
In (b), we see that when the underlying stem is vowel-final, the vowel is deleted in the nominative
form. Now consider (c):
c. CV-loss from stem
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘husband’
// [] [-] [-]
‘meat’
// [] [-] [-]
‘termite’
// [] [-] [-]
‘mullah’
(c) illustrates cases in which an underlying final CV-sequence is deleted in the nominative. The
next example of truncation is shown in (d):
3
d. CCV-loss from stem
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘wooden axe’
// [] [-] [-]
‘dragonfly’
In (d), we see cases in which an underlying final CCV-sequence is deleted in the nominative. Now
examine (e), in which no truncation occurs:
e. Cases with no truncation
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘inside’
// [] [-] [-]
‘sea’
In (e) we encounter cases in which we might expect a stem-final vowel to be deleted in the
nominative (as first seen in (b) above), but in which truncation does not take place; instead, the
stem-final vowel actually surfaces in the nominative form in (e). Now consider (f), in which
augmentation occurs:
f. Augmentation of short stems by vowel epenthesis
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘fish’
// [] [-] [-]
‘head’
In (f), we see cases in which we might expect a stem-final consonant to be deleted in the
nominative (as seen first in (a) above), but in which the stem-final consonant actually surfaces in
the nominative form in addition to an epenthetic vowel. This epenthesis is our first example of
augmentation. A further example is provided in (g):
g. Augmentation of short stems by CV-epenthesis
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘hand’
// [] [-] [-]
‘neck’
In (g), we encounter cases where not only does the underlying stem-final consonant surface, but
we also find an epenthetic CV-sequence in the nominative. A final case of augmentation is
presented in (h):
4
h. More cases of augmentation of short stems by CV-epenthesis
Underlying stem Nominative Nonfuture accusative Future accusative Gloss
// [] [-] [-]
‘speech’
// [] [-] [-]
‘some’
Finally, in (h), we are presented with cases which resemble those in (g) but in which the epenthetic
consonant is different.
Tasks
Your job is to develop an optimality-theoretic account of the Lardil data. Your focus should be
on the nominative forms only; please do not worry about accounting for the other forms. They
are provided solely to give a complete paradigm, but your analysis does not need to extend to them
here.
Approach this problem step-by-step. To assist you in this, analyze one set of data at a time,
and address the following questions, but keep in mind that your final exam may not be presented
simply as a list of answers to these questions. Rather, you must submit a clear and concise
document written in complete sentences, and you must include examples of data and tableaux to
illustrate each step of your analysis.
(1) What do the data in (a) tell you about syllable structure in Lardil? Does the language permit
codas? If so, does it place any restrictions on codas? With this information, you should be
able to determine the relevant markedness constraint to handle the data in (a). Do
underlying stems surface faithfully in their nominative form? The answer to this question
should lead you to pick the correct faithfulness constraint at issue here. How is this
constraint ranked with respect to the markedness constraint regarding codas? Demonstrate
this ranking in a tableau that considers at least two candidates: the optimal candidate, and
the faithful candidate. Your tableau must be accompanied by clear, concise explanation in
complete sentences. Also, are there other potential candidates to consider? If so, show how
they are ruled out. (4 points)
(2) The data in (b) are admittedly weird. To help you account for these forms, I am providing
you with the following constraint – use it in your analysis:
FREE-VOWEL
“An underlying vowel at the end of a stem must be deleted in the nominative form on the
surface.”
First of all, consider the constraint FREE-VOWEL: is it a natural-seeming constraint, in your
view? If so, why? If not, why not? Write at least three clear, concise sentences discussing
your opinion of this constraint, and include arguments and evidence for your views.
Regardless of your answer here, please make sure that your analysis uses this constraint.
As the data in (b) demonstrate, the constraint FREE-VOWEL must outrank (at least) one
important faithfulness constraint. Which faithfulness constraint is this? Illustrate this
5
crucial ranking with a tableau that considers the faithful candidate and the optimal
candidate, as well as any other potential candidates. (2 points)
(3) Now consider the data in (c). Account for these data using clear, concise explanation in
complete sentences. Are any additional constraints necessary, or are the constraints from
questions (1) and (2) sufficient? Demonstrate how the forms in (c) are accounted for with
at least one tableau and a concise, clearly written explanation accompanying it, though
more than one tableau may be useful. (2 points)
(4) Now analyze the forms in (d). Again, do you need any new constraints? If so, what are
they? Illustrate the analysis of forms in (d) with at least one tableau and a concise, clearly
written explanation accompanying it. (2 points)
(5) Examine the data in (e). Given what you’ve seen in (a)-(d), you should expect the forms in
(e) to behave differently from the way they actually behave. Describe this difference in
behavior in plain English using clear, concise sentences. In order to account for the forms
in (e), you’ll need another markedness constraint, which I will provide you with here – use
it:
MINIMALWORD
“Words must be at least two syllables long on the surface.”
First, consider the constraint MINIMALWORD: is it a natural-seeming constraint in your
view? If so, why? If not, why not? Have we seen any other data this semester where this
constraint might have played a role? Write at least three clear, concise sentences discussing
this issue, and include arguments and evidence for your views. Next, discuss how the
constraint MINIMALWORD must be ranked in Lardil with respect to the constraint FREEVOWEL.
Illustrate this ranking with at least one tableau, though more may be useful. Your
tableau must be accompanied by clear, concise explanation in complete sentences. (2
points)
(6) Now look at the forms in (f). Given the constraint MINIMALWORD, do you expect the
underlying stems in (f) to surface faithfully in the nominative, or do you expect to find
what you see there? Describe how the nominative forms in (f) are unfaithful to their
underlying stems. Which constraint is being violated? How must this constraint be ranked
with respect to the other constraints? Illustrate each crucial ranking with at least one
tableau, though more may be useful, and accompany each ranking with clear, concise
explanation. (2 points)
(7) (g) and (h) present an even more complicated scenario: not only do the nominative forms
contain more material than the underlying stems (similar to what you found in (e)), but
they seem to contain an excess of epenthetic material. Think about why this excess material
might be present. What constraint does having all this material satisfy? (Hint: it’s an
Alignment constraint!) Illustrate the analysis with at least one tableau for each crucial
constraint ranking that you require. Accompany each ranking with clear, concise
explanation. (4 points)
6
(8) Finally, consider the forms in (h) and describe in clear, concise prose what differentiates
these from the forms in (g). How are these forms accounted for? Illustrate the analysis with
one tableau for each crucial constraint ranking. Accompany each ranking with clear,
concise explanation. (2 points)
(Extra credit problem starts on the next page)
