Phonology: case studies
By Scott Myers and Megan Crowhurst
Department of Linguistics at the University of Texas
How nasal assimilation and cluster simplification work together
The rules of Cluster Simplification and Nasal Place Assimilation interact in interesting ways that at first glance seem to be contradictory. We know that two rules interact when properties of surface forms can only be explained if the rules apply in a particular order. Ordered rules can stand in what is called a feeding relation, or alternatively, in a bleeding relation to one another. Feeding and bleeding are defined in (9).
| (9) | Interactions between ordered rules. |
| Given two rules A and B which apply in that order (A first): |
| a. Feeding. We say that A feeds B if the result of applying A creates the environment that B needs in order to apply. |
| b. Bleeding. We say that A bleeds B if the prior application of A robs B of the environment B needs in order to apply. |
We will illustrate these concepts using as an example the sentence in (10). (Note that throughout this discussion, the symbol '+' is used to indicate a morpheme boundary, and # to indicate a boundary between words.)
| (10) | Tinc vim botas 'I have twenty wineskins' |
| /tin + k # bint # bota + s/ ➝ [tiŋ # bim # botə + s] | |
| have-1sg twenty wineskin-plural |
Let's first consider the phrase vint botas 'twenty wineskins', which has the pronunciation [bim botəs]. The UR for [bim] is /bint/, with an alveolar cluster. The form [bim] can be explained only if the rule of Cluster Simplification applies first to delete /t/, followed by Nasal Place Assimilation, which takes the input /nb/ and yields [mb]. The derivation in (11) shows how this analysis works.
| (11) | Cluster Simplification must precede Nasal Place Assimilation. |
| UR | |
| Cluster Simplification | |
| Nasal Place Assimilation | |
| Surface form | '20 wineskins' |
The counterderivation in (12) shows that applying these rules in the opposite order does not produce the correct result for the phrase vint botas.
| (12) | Applying Nasal Nasal Place Assimilation before Cluster Simplification produces the wrong result. |
| UR | |
| Nasal Place Assimilation | |
| Cluster Simplification | |
| Surface form |
When we consider the rest of the sentence Tinc vim botas, we see an apparent ordering paradox. While we motivated the rule ordering in (11) based on the phrase vim botas, the first part of the sentence, tinc vim seems to require the opposite order: in this case, the nasal in tinc has assimilated to the velar stop, yielding [ŋ], whereas the ordering in (11) would predict [m] as the result of assimilating to [b]. The derivation in (13) exposes the problem.
| (13) | Must Cluster Simplification really precede Nasal Place Assimilation? A problem. |
| UR | |
| Cluster Simplification | |
| Nasal Place Assimilation | |
| Surface form | wineskins 'I have 20' |
The correct outcome [tiŋ bim] seems to require that Nasal Place Assimilation precede Cluster Simplification, but this is the ordering that produces the wrong result for vint botas (see (12))! (The counterordering of these two rules incorrectly predicts *[tiŋ # bin # botə + s].)
The solution proposed by Mascaró (1978) appealed to the notion of the phonological cycle. Mascaró noted that an important difference between the cluster /n+k/ in tinc and /nt/ in bint is that /n+k/ is formed across a morphosyntactic boundary, that between a verb root and suffix, whereas /nt/ is not. Also, when Cluster Simplification deletes /t/ in bint, a third cluster /n#b/ is formed across another morphosyntactic boundary (a word boundary, indicated by #).
| (14) | A cyclic analysis of Nasal Place Assimilation and Cluster Simplification. |
| Root cycle (Cycle 1) |
Input to Cycle 1 | |||
| Nasal Place Assim. | ||||
| Cluster Simplification | ||||
| V reduction | ||||
| Output of Cycle 1 | ||||
| Cycle 2 (Affix) |
Input to Cycle 2 | |||
| Nasal Place Assim. | ||||
| Cluster Simplification | ||||
| V reduction | ||||
| Output of Cycle 2 |
| (15) | Postlexical Cycle (phrase level). |
| Postlexical cycle |
Input to cycle | |
| Nasal Place Assim. | ||
| Cluster Simplification | ||
| V reduction | ||
| Surface form |
We've now come to the end of our examination of the processes Nasal Place Assimilation and Cluster simplification in Central Catalan. We have provided a generative derivational analysis, in which the effects seen in the data are attributed to phonological rules that apply to series of abstract representations. A different kind of theoretical analysis would use the tools provided by the theoretical framework Optimality Theory. For an OT account of the processes of nasal assimilation and cluster simplification we have seen in this section, we recommend the paper cited below.
Reference
Herrick, Dylan. 2002. 'Catalan Phonology: Cluster simplification and nasal place assimilation.' In Joaquin Camps and Caroline Wiltshire (Eds.), Romance Phonology and Variation: Selected papers from the 30th Linguistic Symposium on Romance Languages. John Benjamins, The Netherlands. pp. 69-84.