For well over a hundred years, philosophers of language have been asking questions that are abstract and general – in the extreme: How is it that, with language, we communicate thoughts to one another that are objective, about things in the world itself, not just something about our own ideas? What is the relationship between the concepts we think with and the meanings of words? How can the thoughts we express in language be true or false? These foundational questions are about the meanings of words, how the mind represents the world, and how, with words, we refer to aspects of the world and accurately represents the way the world is.
I say “Amanda Gorman is from Los Angeles.” It seems that I refer to a particular person, the extraordinary poet who spoke at Biden’s inauguration. When you hear what I say, you understand something about Amanda Gorman herself. But how can that be? I have my own private subjective ideas of Gorman and you have yours . . . aren’t we each trapped in our own minds, limited by our own sets of knowledge and experiences? Language – the meanings of words – is supposed to be the interpersonal glue that rescues us from solipsism, hooking us onto the shared “external” world.
We can ask similar questions about natural kind terms like “water.” What does “water” mean? Is it synonymous with a description like “clear potable liquid”? Probably not. Because if it was, when I speak of water, I might well be speaking about something that only looks like water, but isn’t. Then does “water” have the same meaning as “H2O”? We successfully communicated about water long before the discovery that water is H2O – and the discovery didn’t ring in a new meaning of “water.” So, again, how is it that with the word “water” we can and do speak about that specific precious liquid of planet Earth?
Gottlob Frege, the founding father of semantics, lay out these questions (and more) in his ground-breaking 1892 essay “On Sense and Reference.” For over a hundred years, philosophers of language took Frege’s problems as agenda-setting. And with good reason: the puzzles and questions Frege laid out are deep, and his insights into the interconnectedness of meaning, reference, concepts, truth, and logic advanced our understanding of language far beyond that of any thinker before him.
Yet Frege had enormous blind spots and severely circumscribed interests. Although he did recognise emotive and expressive aspects of language, he saw no value in understanding them. Frege offered “dog” and “cur” as examples of words with the same sense but different tone. “Dog” is neutral, while “cur” has negative valence, indicating the speaker’s negative evaluation or emotion. Frege urged – pretty much without argument – that in a scientific approach to language, tone is irrelevant. So he neglected theorising about expressive words and the social and political dimensions of meaning, and anything that we now understand as sociolinguistics. That’s not surprising. Frege developed his semantic theory incidentally, as a step toward (trying to) prove that arithmetical knowledge is essentially logical knowledge – and so focused dominantly on how language encodes truths.
What is surprising, though, is that in the hundred-plus years since “On Sense and Reference,” while philosophers of language developed “pure” semantic theories of names, kind terms, descriptions, belief attributions, truth, modal discourse, and much more, they shirked analysis of the emotive and social-political dimensions of meaning, including propaganda, coded speech, racist, homophobic, and gendered discourse. And remarkably, they neglected these areas despite enormous attention to them throughout the twentieth century within political theory, sociology, law, and Black, Latino, LBGTQ+, and feminist studies.
We’ve seen a marked shift in the last twenty years. Why has philosophy of language changed so rapidly? Why now?
A confluence of forces propelled the rapid change. I can only touch on a few here. Some stem from philosophical developments within “pure” philosophy of language; others from the acceleration in society’s – and so too the profession’s – intense social and political focus. A third force stems from dramatic changes in the culture of the profession.
Two developments within “pure” philosophy of language stand out for me. One of the central ideas emerging from 70s and 80s “pure” philosophy of language is that reference is not an individualistic, but rather a social affair. We can refer in part because we’re related to others in our community who are experts or otherwise “in the know.” This is Putnam’s idea of the linguistic division of labour. To refer to elm trees, I don’t have to know distinguishing features of elm trees. Instead, I – we –implicitly lean on experts whose specialised knowledge secures reference. Our thoughts and beliefs are similarly socially mediated. I can be wildly off on what arthritis is, thinking I can get it in my thigh, say, yet still have beliefs about arthritis. Amazingly, the classical examples are all deliberately mundane, apolitical words (“water,” “elm tree”) and concepts (arthritis, carburettor, sonata, sofa).
A core feature of the non-individualistic view of reference is that whole communities – everyone – could be wrong about the concepts they think with. Our linguistic norms and total prior practices can be challenged. Why is this so politically important? Think about the fight over gay marriage. In the political sphere, you often hear that gay marriage is impossible (a contradiction in terms!) because what marriage is – what we’d always meant by “marriage” – is a union between a man and woman. That’s wrong. But even if it were true, it doesn’t tell against gay marriage. This research reveals why: what marriage is need not be, and is not, exclusively determined by our community’s prior practices. It depends in part on the social norms we commit to.
This advance in pure philosophy of language spawned the new, vital, socially transformative area of applied philosophy of language known as conceptual engineering. Conceptual engineers craft analyses of socially loaded categories like marriage, gender, race, rape, among others. Relying on empirical, analytical, but also moral and political considerations, they aim to engineer concepts at the foundations of a more just society. So they ask not only “what have we meant by ‘marriage’” but also “what should we mean by it?” And they answer with normative theorising that is – rightly, in my view — unapologetically political.
A second development of “pure” philosophy of language contributed to an explosion of research steering toward the social-political. One of the hottest topics in the late 70s through the 90s was the semantics of demonstratives, expressions like “that” and “this,” and indexicals, expression like “I,” “this,” “here,” “now.” These expressions shift their reference with the context. If I say “I live in Los Angeles,” I refer to myself with “I.” If you say the same sentence you refer to yourself, not me. Frege noticed their special properties, but his semantic theory gets them wrong. In pioneering research, John Perry and David Kaplan demonstrated that demonstratives and indexicals incorporate at least two semantic factors, the character, the “rule of use” that governs the expression (for example, the rule for “I”: use “I” to refer to the speaker of the utterance), and the content, what the expression referred to on an occasion of use. Utterly brilliant work. And as pure as can be.
Remarkably, the pure directly inspired and paved the way to the political. In his dazzlingly original mid-90s paper, “The Meaning of ‘Ouch’ and ‘Oops’,” Kaplan applied his idea that demonstratives are governed by “rules of use” to expressions that colour communication: expressives like “damn,” “bastard,” and “hooray!,” terms of endearment like “sweetheart,” terms of hate and ill-will including slurs and epithets, nicknames of all stripes, honorifics like “Professor,” and all kinds of politically correct and incorrect speech. Although he gave an untechnical treatment, he sketched how to treat these expressions in, as he said, a “serious scientific way.” Semantic analyses of thick terms (expressions like “courageous” and “cruel,” “loyal” and “lewd,” that mix description and evaluation) and semantic accounts of expressivism within metaethics provided another rich source and inspiration for wider investigations within philosophy of language proper.
Starting in the 2000s, a handful of linguists and philosophers of language inspired by Kaplan turned their attention to emotive language and derogatory expressions, especially slurring terms for socially significant groups based on race, nationality, religion, ethnicity, sexual orientation, sexual identification, class, and so on. What’s fascinating about these expressions is how they challenged not only Frege’s disregard for tone but other classical assumptions underwriting how philosophy of language ought to be done.
Think of a slur and compare it with a “neutral” or respectful word for the group. In one way, the difference between the slur and the neutral word seems similar to the difference between “cur” and “dog.” But for the slurs, it’s far more complex. Unlike “cur” and all-purpose put downs like “jerk” and “asshole,” slurs are socially-steeped terms. To understand them, we need to theorise on how language encodes facts about racism, sexism, ableism, etc; how words can make salient or encode group stereotypes; and how particular words can communicate specific vicious emotions; how words can shame and be badges of shame; and how language contributes to the formation of hierarchies. Previous work in philosophy of language excluded these complex – and, yes, messy – social aspects of meaning. These days, they are centre stage. People are asking: How can a word encode, or a use of a word manifest, something so multi-faceted as a racist ideology? Should it be treated on analogy with other co-referencing terms with differing register, like the medic’s request for a sample of feces, a parent’s urging a child to poop in the potty, and the unlucky pedestrian’s curse after stepping in crap? Do certain words function to encode specific emotions? Or do we need entirely new resources – new theories of communication – to explain them?
Of course, forces outside of philosophy of language also impelled change. It’s ironic, really, how much Trump contributed to shifting the scene in philosophy of language. Although the trend toward the social-political has been brewing for two decades, it’s accelerated massively in the last five years. How could it not? Since he announced his presidential candidacy six years ago, we would wake up to headlines about him referring to Japanese people as “The Japs,” calling Mexicans “rapists,” boasting how he grabs women “by the pussy” (and brushing it off as “locker-room talk”), calling the coronavirus the “Kung-flu,” and on and on. Every newspaper in the country would run op-eds informally expounding on how such words poison our social world, dehumanize groups, and ignite violence against them. Philosophers of language – equipped with semantic and pragmatics theories – quickly got in on the game to deepen our understanding of the linguistic, psychological, and social features of his acts.
The worst for me – and an incident I’m currently writing about – was when Trump publicly mocked the New York Times reporter Serge Kovaleski for his arthrogryposis condition during a political rally. He didn’t use words like “gimp” or “crip”; but he performed the very same kind of speech act by flailing his arms, jerking his wrists, dehumanising Kovaleski for his disability. I think acts of this kind whose main aim is to humiliate structurally parallel the moral wrong in torture.
Changes in the culture of our profession – toward making it more inclusive, more diverse, and less confrontational – have also contributed to encouraging investigations in social-political philosophy of language. Philosophy has always been hierarchical and status-heavy: the more technical, the more abstract, the more pure your work, the greater your street-cred. That vibe has long permeated philosophy of language, encouraging following closely in Frege’s footsteps – after all, analysis of racist, sexist, or misogynistic language was previously almost universally written off as trivial, soft, for lightweights only. It was hubris, and, fortunately, that attitude is beginning to wane.
I’m excited to see how in the upcoming decades these trends toward social and political language will impact foundational areas in philosophy of language. Work on socially loaded expressions has already infused new life into foundational research projects in semantics. Prior work investigated meanings statically, by conceiving meaning as fixed to a particular place and time. Recent work on terms like “marriage,” “rape,” “woman,”takes the lexicon itself as dynamic. Prior work didn’t examine the mechanisms by which the meanings of words change over time. Partly inspired by conceptual engineering, partly inspired by understanding slur reclamation like taking back “queer” as a respectable identity label, new work investigates the processes by which words change their meaning over time. Previously, philosophers of language largely ignored the fascinating subject of polysemy – how words can have multiple related meanings. I’m confident the trend toward the social-political will itself inspire richer and more general accounts of language and communication.