e1. what is linguistics even modeling?
how my meaning gets in your head, and an introduction to this blog
People become linguists for different reasons. Some people want to help save indigenous languages. Other people are fascinated by cross-linguistic diversity: the huge variety of sounds, kinship systems, and morphosyntactic strategies for imparting different types of meaning. Still others are interested in language acquisition in young children, or in wannabe-bilinguals. Maybe they watched Arrival. Maybe now some people want to become linguists to understand ChatGPT.
Somehow a lot of us end up drawing trees. Syntax has trees, phonology has trees, semantics has trees, computational linguistics has trees with math… the list goes on. The reason for this is that language is hierarchical on many levels, as we’ll talk about later, but for now…
Drawing trees is really fun, but sometimes it feels like you’re not really doing anything meaningful with your life. Maybe the government threatens to take away your grad school opportunities because you’re not actively contributing to society. Maybe your friends and relatives ask you to explain yourself and… what are you doing, exactly, with those trees? Linguistics is a science, and science is about creating accurate models of the world around us: biologists draw cells, physicists draw trains, and astronomers draw stars. So why do linguists draw trees?
To the best of my knowledge, theoretical linguistics is supposed to be modeling whatever it is that humans have in our head that allows us to talk to each other. That’s what I’ll be exploring in this blog—from now on I’ll just call it “language.”
It might seem unintuitive that humans actually have something in our brains that corresponds to language. Even more unintuitive: in our genes. How would that even work?
I mean… I don’t really know. That’s the whole point of this blog.
There are a lot of arguments for why we have something in our genes that corresponds to language1 (the poverty of the stimulus is probably the most famous one), but my personal favorite is this: why don’t other species have language?
Even if there is another species that has a language just as rich as ours, and we simply have yet to discover it, it’ll still be the case that the vast majority of other species cannot talk like us. Including our closest relatives.
But I’ll talk more about animal communication and the evolution of language next time—what’s important now is that there’s something in our genes,2 which puts something in our brains, which is the ability to learn and use language.
how i get something from my head to yours
If I want you to know that my mom likes chocolate, the following things need to happen (absent of you seeing my mother eat a bunch of chocolate chips):
A meaning exists in my head. We’ll model this meaning with its logical form:
likes(x, chocolate) & mom(me, x). A logical form is a translation of a meaning into some form of predicate logic. We’ll worry about the exact form later.I search through the lexicon of my language, figuring out which words are necessary to express this logical form. In English these words are “like,” “mom,” and “chocolate,” as well as some form of “me.” Then I put these words into an order in my head, and inflect them in various ways. For some reason I settle on “
my mom likes chocolate” as the order and inflection that represents the logical form above.I figure out the phonological forms of each word, and say them one after another:3
[mäjˈmämlɐjksˈtʃɔkɫɪʔ].You hear
[mäjˈmämlɐjksˈtʃɔkɫɪʔ], and break it up into a series of distinct phonological ‘words,’ which you then put into a sentence by matching them to the phonological forms of words in your lexicon:my mom likes chocolate.You somehow get from that sequence of words to the meaning that I intended, in logical form:
likes(x, chocolate) & mom(me, x).
semantics, syntax, phonology
Speaking language and understanding language are inverse processes. When speaking, you have to transfer a logical form (LF) to a sequence of words, and then to a final sonic signal, or phonetic form (PF). When listening, you have to transfer a phonetic form to a sequence of words, and then to a final logical form. Semantics basically deals with LF, syntax with sequences of words, and phonology with PF. Note that when I use the words ‘semantics,’ ‘syntax,’ and ‘phonology’ in this blog, I will sometimes be referring to the cognitive modules that process logical forms, sequences of words, and phonetic forms respectively, and other times be referring to the study of these modules.
Going from sequences of words to phonetic forms and back again doesn't seem too difficult. We definitely don’t know exactly how sequences of words become sound waves, but phonologists and phoneticians have really been Doing The Work for the past century, and so we have a pretty good idea.
Going from logical forms to sequences of words and back again is much harder from a theoretical standpoint. Getting from my mom likes chocolate to likes(x, chocolate) ∧ mom(me, x) and back again. Now this is difficult. This is what syntax and compositional semantics are all about. (Even phonologists are invited to this party, using things like pitch contour and contrastive stress to explain how we disambiguate possible meanings.)
As we’ll see in the next post, some animals have both conceptual structure (which can be modeled as some sort of predicate logic) and phonology. But no known non-human species has the ability to move things from conceptual structure to sequence of words to sonic signal. Only humans.
So how do we do it?? Most crucially, how do we get from a meaning to a sentence and back again?
What is it in our brains that allows us to do this, and restricts the abilities of other species to do it—even when they have phonology and conceptual structure?
This is The First Big Question. This is what this blog is going to be about.
what this blog is going to be
There’s a lot of stuff I don’t know about linguistics. But I’m a firm believer that figuring things out yourself, making your own personal theories and hypotheses about things, and then reading up on what other people think, is always better than the reverse.
That said, reading about what other people think is also extremely important. Especially for me, given that I’m an undergrad and these questions have been pondered by adult linguists for at the very least 60 years. It’s likely that I will say things that are completely wrong, and that I will partially re-derive existing theories without crediting them. (Call me out on these things!)
But anyway, I will be reading a bunch of literature at the same time as writing this blog, and because of this, my theories will probably change at certain points along the way, reflecting new insights I’ve gained. But I’ll always try to build the theories here from the ground up. Always tying them back to The First Big Question (and the Second, which I’ll introduce in the next post).
I think often I have so much fun playing around in the weeds of my trees that I forget what they’re actually representing, and why I’m drawing them in the first place. I take the extremely specific theories around me for granted. That’s why this blog is going to be a little bit heretical. A little bit funky. I am almost definitely going to say things that are not part of current mainstream linguistic theory, sometimes even on purpose, and that’s part of the point!
Linguistics grad student STM once told me that belief in syntax follows a bell curve: at first you think it’s real, and then you think it’s fake, and then you think it’s real again. I want to Believe In Syntax, and this blog is a way for me to make sure I really understand the motivations behind everything.
The posts on this blog will probably be a mixture of things, but they're sure to include the following:
posts about animal communication
posts about human evolution, genetics, and neuroscience
posts about information theory, computer science, and LLMs
(perhaps most frequently) posts about syntactic and semantic formalisms
I've already written ~12 posts. Check out the table of contents here.
why i’m sharing these thoughts with the world
I’m a firm believer that blogs are just complicated ways to find cool people and get them to send you interesting things. So please, I implore you—comment! Reach out to me and tell me why I’m wrong! Tell me what theory I’m missing, or what cross-linguistic data needs to be explained! Send me papers and constructions and any philosophical or theoretical musings.
Writing things out is a also great way to understand somewhat-fuzzy concepts in your head. In trying to explain something to someone, you’re kind of also explaining it to yourself. At least, this works well for me.
Lastly, I’ve been really wishing there was a theoretical linguistics blog or podcast I could read or listen to, and I haven’t found one. Theoretical linguistics is really interesting, and I wish more people understand what it was actually about. Forget about me not believing in syntax—the people around me? They have no idea what we’re doing or why it’s important! I want my friends and family to understand and appreciate the work that us linguists are doing.4
who this blog is for
I will try to write everything out in a way that is both (1) understandable for people who are not theoretical linguists and (2) not boring for theoretical linguists. However, this does not mean this blog will be ‘introductory.’ I will dive straight into some pretty deep theory, and it is up to you, my dear reader, to read carefully.
This blog is not a summary of current linguistic theory, although you will probably learn a lot about current linguistic theory if you don’t know it already. The theories I present in this blog, unless I tell you otherwise, are my own. They will thus be almost inevitably flawed, but that’s kind of part of the point. I will also probably not cite enough, because most of what I present here will be emerging from my messy brain that does not keep track of where all of its ideas come from. Sorry!
(If any of my professors ever read this—I apologize in advance for butchering what you’ve taught me! But also, thank you :)
This blog is for anyone interested in linguistics, animal communication, computer science, music, evolutionary biology, and cognitive science. Maybe also psychology, philosophy, and neuroscience. I will be trying to figure out what humans have in our head that corresponds to our ability to learn and use language, how it evolved, and what parts of it are shared with other species.5
Here goes :)
to be clear: that corresponds to our ability to learn and use language productively. not specific languages like English or Arabic—those are definitely learned from our environment.
importantly, the thing in our genes doesn't have to only be used for language, or only be possessed by humans, although many people think it is. complex language could theoretically also be an emergent phenomenon.
with some variation in pitch, time, and timbre. Also, these symbols are from the International Phonetic Alphabet.
To be clear, I do not think ‘theoretical’ linguistics is any more important than other types of linguistics. Documenting and revitalizing indigenous languages, for example, is extremely important. They are just different types of important, different areas for growth. But also, know that this is not a full dichotomy! A lot of really great linguists do both theoretical linguistics (trying to figure out UG and formalisms and stuff) and language revitalization, computational linguistics, psycholinguistics, etc.