From my perspective, Pater's (2019) target article does a great service both to researchers who work in generative linguistics and to researchers who utilize neural networks—and especially to researchers who might find themselves wanting to do both by harnessing the insights of each tradition. The fusion of theories of linguistic representation and probabilistic learning techniques has certainly led to many interesting and valuable insights about the nature of both linguistic representation and the language acquisition process. However, I feel that the most exciting aspect of Pater's article is the increasing interpretability of neural network models, especially when combined with insights from the theoretical framework of generative linguistics. This allows for the possibility that neural networks could be used to actually generate new theories of representation. I describe how I think this theory-generation process might work with interpretable neural networks.

Pater's (2019) target article builds a persuasive case for establishing stronger ties between theoretical linguistics and connectionism (deep learning). This commentary extends his arguments to semantics, focusing in particular on issues of learning, compositionality, and lexical meaning.

Pater's (2019) target article proposes that neural networks will provide theories of learning that generative grammar lacks. We argue that his enthusiasm is premature since the biases of neural networks are largely unknown, and he disregards decades of work on machine learning and learnability. Learning biases form a two-way street: all learners have biases, and those biases constrain the space of learnable grammars in mathematically measurable ways. Analytical methods from the related fields of computational learning theory and grammatical inference allow one to study language learning, neural networks, and linguistics at an appropriate level of abstraction. The only way to satisfy our hunger and to make progress on the science of language learning is to confront these core issues directly.

Whether lexical representations are stored as abstract forms or exemplar tokens is the focus of much debate in both the phonological and word-recognition literature. This research report examines the recognition of words that have undergone Canadian raising and/or intervocalic flapping. Two eye-tracking experiments suggest that listeners are slower to fixate words that have undergone one or more phonological processes within their own raising dialect, supporting the idea that they must calculate a mapping from surface word forms to more abstract representations. Implications for representational and phonological theories are discussed.

It has become increasingly clear that current threats to global linguistic diversity are not restricted to the loss of spoken languages. Signed languages are vulnerable to familiar patterns of language shift and the global spread of a few influential languages. But the ecologies of signed languages are also affected by genetics, social attitudes toward deafness, educational and public health policies, and a widespread modality chauvinism that views spoken languages as inherently superior or more desirable. This research report reviews what is known about sign language vitality and endangerment globally, and considers the responses from communities, governments, and linguists.

It is striking how little attention has been paid to sign language vitality, endangerment, and revitalization, even as research on signed languages has occupied an increasingly prominent position in linguistic theory. It is time for linguists from a broader range of backgrounds to consider the causes, consequences, and appropriate responses to current threats to sign language diversity. In doing so, we must articulate more clearly the value of this diversity to the field of linguistics and the responsibilities the field has toward preserving it.