6

Certainly! Humans can do this too; computers are just more consistent at it. In general, language usage doesn't just come down to what's grammatical and what's not. A man or a woman can say "I like apples" and have it be equally correct. The differences come in when there are multiple ways to express the same concept. For example, Jespersen(1) conjectured ...


5

We simply do not know: this is an empirical question, and nobody has done the study. The problem with writing in IPA is that you have to understand what the various letters "mean". It's easy for Saami speakers to learn that the letter č "means" (in IPA) the phoneme /tʃʰ/ and ž "means" the phoneme /tʃ/, and the don't need to worry about low-level details. In ...


3

This is a very difficult question to answer! And the best I can offer is, right now, we don't really know—these mental structures are still very much a mystery! are the closer signals closer in sound (phonetically) or in meaning (semantically)? "Closer" is hard to define. But one way to study this is priming. Basically, if you show people a series of ...


3

What you remember is quite vague, but I think this is related to word vectors. Word vectors are internal representations of words learned by a neural network, and they live in a high dimensional Euclidean space (typically several hundred dimensions). One algorithm to get at word vectors is word2vec. It is still poorly understood why the word vectors have ...


3

Any studies of reading fluency (e.g., can people read fraktur better than antiqua?) boil down to the trivial: One reads best what one is used to read. It is also known that we skip a lot of information encoded in the writing, e.g., we don't notice a lot of typos or printing errors because the redundancy in the written language is that large. So, given a ...


2

There's this, but it wasn't fMRI. It's direct cortical recording! http://science.sciencemag.org/content/early/2014/01/29/science.1245994.abstract


2

There was a paper by Huth and colleagues in 2016 which seemed to somewhat reinvigorate interest in a detailed semantic map of the brain: https://neuroscience.berkeley.edu/detailed-map-language-representation-human-brain/ More recently, the same group at Berkeley (or at least sharing some members) have shown the validity of their semantic maps in a modality-...


1

One can use the cosine similarity measure in word space. Google up some papers by Mikolov, who describes how the models can be generated. It’s quite remarkable that they can derive, for example, |king⟩+|woman⟩-|man⟩≃|queen⟩.


1

Neuroscientists/neurolinguists don't typically use the term "Merge" (which comes with some theoretical baggage that they may or may not ascribe to), but there's definitely been some recent neuro work on combinatory linguistic operations. David Poeppel and Liina Pylkkaänen are researchers who come to mind. Some quick googling turned up the following: http://...


Only top voted, non community-wiki answers of a minimum length are eligible