Reflection

Developing my problem set

It was relatively straightforward to create the first draft of the problem set. The step-by-step instructions were easy to follow. I tried to keep the phonotactics of the language simple, since this is a morphology class, not a phonology class! All of the roots have a CCVC structure, except gròwt ‘person’, which has a diphthong. I used fairly simple phonemes and made sure the roots were all quite different. I made one of my roots, bjál ‘to be pretty’ a verb, even though it is an adjective in most languages I am familiar with. I used a mix of prefixes and suffixes for my derivational affixes, to keep it interesting, but kept all of the inflectional morphemes as suffixes, as that is more common according to WALS (Dryer 2013).

For my derivational affixes, I used an agentive and an augmentive. I created an agentive prefix fì-, similar to English -er, but I used it to form words with idiosyncratic meanings, which such as fì-dràd, which literally means ‘growing one’ but I glossed it as ‘child.’ I also created an augmentative suffix, inspired by the diminutive (but the opposite). Augmentatives do exist in some languages (for example Cinque 2015: 70 includes some Italian examples), although they seem to be less common than diminutives. I combined both the augmentative and the agentive in a single word to create fí-swív-on ‘prophet.’ I didn’t include fí-swív ‘seer’ in my data set, but I was thinking that fí-swív would be anyone who sees, while adding the augmentative creates a special kind of seer. I considered including fì-dràd-on to mean ‘teenager’ but I thought the problem set was getting too long so I left it out. I also made sure to include some forms with less idiosyncratic meaning to make it clear what the function of the derivational morphemes were.

I had both verb roots and noun roots in my original list, but I included a nominalizing derivational prefix, which meant that all my roots could appear in noun form, so I decided to make a nominal inflectional paradigm. I decided to distinguish case, number, and animacy. Case and number are quite common, so I decided to add animacy, which is less common, to keep things interesting. The animacy distinction is inspired by the Algonquian language family, as discussed in the textbook section on noun inflection. I opted for a blend of agglutination and fusion, and I fused case and animacy into a single morpheme, which makes the problem set trickier! I thought about including a dual number or more case values, but I thought the paradigm was getting too complex, so I left them off.

All of the morphemes included so far were concatenative, so to add a non-concatenative morpheme, I added tense marked as tone on the verb root.

Finally, I added a -r/re and -n/ne allomorphic pattern to the number suffixes based on whether it was sentence final. I think this is an unusual allomorphic pattern, but I did learn about one like this once in a phonology class, where there was a special morpheme that occurred clause-finally (CITATION). I made sure there were both transitive and intransitive sentences so that it would be clear that it was about being sentence-final rather than a case marking.

Exchanging problem sets

I exchanged problem sets with Chris and Melissa.

In the first version of Chris’ problem set, he only had one word showing each morpheme in the data set, so it wasn’t possible to see the patterns. I didn’t even know where to start! He revised it and showed me again, and it was a lot better. However, even in the revised version, there was still not quite enough data. One word was glossed simply as ‘burned’, and it wasn’t clear whether the word was supposed to be an adjective or a past tense verb, since the English word in the gloss is ambiguous. It had a past tense marker, but that was the only word in the data set where the allomorphy in the past tense showed up, so I wasn’t able to identify that it had the past tense marker nor figure out the allomorphic pattern. There was also only one word that showed the reduplication pattern, so that made it difficult to tell exactly how reduplication would deal with different word shapes, such as consonant clusters. I realized that the more complex or unexpected a pattern is, the more data someone will need in order to solve the problem set.

Melissa’s problem set was pretty good. One issue was that there was an <l> that appeared on either side of the morpheme boundary, as shown in (14), and I couldn’t tell whether that was mean to be allomorphy or a mistake. 

Comparing (14a) and (14b), it looks like the roots are zul ‘carry’ and drosh ‘break’ with -ta as the past tense marker. But if the root for break is drosh, with no <l>, there should either be <ll> in (14c) or no <l> in (14d). One possible solution is that the agentive suffix is -lekta but when there are two <l>s in a row, one deletes. I asked Melissa, and she said it was a mistake, but that she’s going to keep it in as another allomorphic pattern in her final version! But she added another example to make it clearer.

Melissa also had two words with a na- prefix, one which corresponded to negation and the other to modal ‘might.’ It was interesting that maybe this language would have an underspecified irrealis mood, but I wonder if the difference between negation and ‘might’ needs to be explicit in an actual language in order for people to not be confused.

Revising my problem set

I showed it to Chris first, revised it, and then showed it to Melissa.

Chris said it seemed weird that there were tones on the verbs but not the nouns, so I added tones on the verb roots. I didn’t add tones on the suffixes, because some languages have some neutral tone suffixes. For example, Mandarin’s fifth tone is considered to be a “neutral tone.” The pitch of syllables with the neutral tone is not phonologically specified; instead, it is predictable from the preceding morpheme (Yip 1980: 79-80). Adding tones to the nouns also gave me another opportunity to add allomorphy. I decided that the fi- prefix should match the tone of the following root. I think it was good to give a second allomorphic pattern, since the sentence-final -e could be analyzed as a separate morpheme rather than allomorphy. By adding the tonal allomorphy on fi-, there is now a non-ambiguous instance of allomorphy in the data set.

Chris also suggested changing ‘skyscraper’ to ‘mansion’, but I liked having a more idiosyncratic meaning, and both Chris and Melissa were able to figure it out pretty quickly. However, I decided to change the gloss of the associated root plíg from ‘house’ to ‘house/building’ in the answer key.

Melissa didn’t find all of the morphemes in the solution. She noticed the tones but ignored them and she wondered about case but didn’t actually figure it out. Even though she didn’t get those right, I decided to leave them in because she noticed both patterns. I think she probably would have figured it out if she took a little more time, and I don’t want the problem set to be too easy. However, she completely missed the animacy distinction, so I decided to add a minimal pair for animacy to make that clearer. The animacy distinction is especially difficult to figure out because the morpheme that indicates it is only one vowel long and is fused with case. However, it was tricky to come up with a minimal pair for animacy without also incorporating cultural knowledge (like how sacred plants are sometimes animate in Algonquian languages) or a story (with an anthropomorphic talking plant or something), but eventually I settled on distinguishing people and statues, with the root for statues being the same as the word for giant, but with the inanimate marker. I think adding this miminal pair will also help students find the case markings, because you can’t figure out the distribution of the animacy markings without taking case into account.

References

Cinque, Guglielmo. 2015. Augmentative, pejorative, diminutive, and endearing heads in the extended nominal projection. In Elisa Di Domenico, Cornelia Hamann, and Simona Matteini (eds), Structures, Strategies, and Beyond: Studies in honour of Adriana Belletti. John Benjamins. 67-82.

Dryer, Matthew. 2013. Prefixing vs. suffixing in inflectional morphology. In Matthew Dryer and Martin Haspelmath (eds.), WALS Online. Available online at http://wals.info/chapter/26. Accessed on 2025-04-17.

Yip, Moira. 1980. The tonal phonology of Chinese. PhD thesis, MIT.