I was unfortunately unable to answer the questions in the chat after the lecture and discussion with Edith Moravcsik. Here are attempts to answer those questions: ***Luana Amaral: can you talk a little bit about the class of words we normally call "adverbs"? There are two issues here, I think. The first is the semantic or functional range of concepts described as "adverbs". It is such a huge range, from manner to spatial/temporal/other deixis to modal, speech act, and discourse markers, that "adverb" doesn't really represent a coherent semantic class. In terms of syntactic distribution, "adverbs" vary greatly within a language, let alone across languages. It is better to divide "adverbs" into semantically coherent classes, and then compare them typologically. One of those classes would be manner words like "fast", "slow", "well" and "badly". These are sometimes analyzed as predicate modifiers. I have restricted modification as an information packaging (specifically, propositional act) function to modification of referents (prototypically, adjectives). But Pernilla Hallonsten Halling, in her dissertation and a book forthcoming from Oxford U Press, uses my part of speech analysis to argue for a predicate modifier prototype as a useful typological category (=comparative concept). ***Luana Amaral: this "competition" between different factors, such as human and agency define variation in structures, right? can you talk a little bit about this regarding psych verbs? Psych verbs, or experiential events as they are called by Elisabeth Verhoeven in her 2007 book, are a classic case of events whose participant roles don't fit very well with the combination of semantic role (volition/intention/control) and topicality -- the human experiencer ought to be highly topical by virtue of being human, but it is low in control in most experiential events. So they are quite variable in their case marking across languages -- and, as I discovered, in their verbalization in English, which from a form-to-meaning perspective is "strongly" experiencer-subject. The patterns are more subtle, as I indicated in my lecture. I think there are even some typological patterns here. Events describing an experiencer attending to the stimulus, like "look at", will have the experiencer encoded as Subject. Events describing a stimulus causing a change of mental state in the experiencer, like "scare", will have the stimulus encoded as a Subject (and sometimes also have causative morphology). Events simply describing the experience, which are stative (although they may also be construed as inceptive, like many states), like "see", are highly variable in their argument encoding across languages. ***Marcia dos Santos Machado Vieira: Could you talk about how do you deal with variants and the cross tabulation of their explanatory sociolinguistic factors/predictors when dealing with a cross-linguistic description? Your question brings up a dimension that I didn't have time to cover in this lecture. I looked at distributional variation and particularly variation in verbalization in a single language, and showed the parallels to cross-linguistic (typological) variation and generalizations. But how do you get from one to the other? I have argued that this process is mediated by sociolinguistic factors. First there is variation in verbalization. Then the variation acquires some social valuation, so social factors predict their occurrence (probabilistically again). Then a variant comes to be fixed as the conventional way of expressing the function. But the social factors driving conventionalization are independent of the functional factors that generated the variation in verbalization. They are also specific to the speech community in question. (You can find certain patterns across societies, such as women leading change in a number of societies; but that is still independent of the functional factors.) This is another dimension in the overall theory of language, a theory where variation and change are essential aspects of language. I have discussed this in "Explaining Language Change: An Evolutionary Approach" (2000), plus some follow-on papers. ***Barthe Bloom: I was wondering about some implications on the analysis of language change. Inflectional morpheme developing out of overt marking, then implies a shift to centralization on the prototype, right? Regarding paths of grammaticalization and the semantic change that is involved: I'm not sure that there is a prototype structure to the conceptual spaces that represent semantic change in grammaticalization. Not all conceptual spaces have prototype "points" in them. It has been extremely difficult to come up with some overall semantic parameter governing all types of semantic change that have been called "grammaticalization". I have made four different proposals myself, and I'm not the only one. The last one was about verbalization, based on Chafe's original theory of verbalization. Clearly frequency plays a significant role. But perhaps it is better to try to understand better the paths of change for specific grammatical semantic categories, such as perfect > recent past > general past or direction > opposition > comitative.
