Sense, Reference, and "What is a Woman?"

A context-based understanding of meaning

Language is a window into the mind. While we may not have direct access to the content of each other's thoughts, our power of language allows us to create and understand representations of thoughts. But this seemingly simple ability contains significant complexities.

For one, the same word can express distinct yet related meanings in different contexts—what is known as “polysemy.” Language isn't only about understanding semantic and grammatical rules, but also how those rules fit into context to create meaning.

For example, the word "body" can express different meanings depending on whether one refers to a human, an airplane, a large accumulation of water, an e-mail, a collection of works, or good wine. And all of these possible senses of the word "body" stand for multiple expressions of essentiality, mass, or substance.

There is a general concept of "body" that allows it to have a specific meaning under different sets of circumstances. This general concept is known as the "monosemy" of the term—the abstract idea from which a word’s specific uses derive.

Words don't represent a single, fixed concept. Rather, they can contain a dynamic collection of thoughts. But analytic philosophy fails to appreciate this notion of multiple senses — instead, viewing a word as standing for a single, unchanging concept that fits neatly into a logical structure. Nevertheless, language is not mathematics. 1+1 = 2 is true irrespective of context. The meaning of words, meanwhile, is incredibly context-dependent. 

In this article, I intend to explain the important distinction between the reference of a word and its meaning, as well as how polysemous words like "woman" can convey different yet legitimate meanings in different contexts.

Sense and Reference

To use a famous example, If I say, "Abigal knows that Cicero was a Roman statesman," the knowledge I'm attributing to Abigal is that of Cicero's profession. Yet if I were to say, "Abigal knows that Tully was a statesman," the truth value of the sentence changes. To Abigal, "Tully" may not represent the thought of the famous Roman statesman, so therefore she may not know anything about "Tully's" profession. Even though both attributions of knowledge refer to the same person, as Tully is just another name for the famous Roman statesman Marcus Tullius Cicero, Abigal may not know this about "Tully," which to her is a meaningless word. Since some people do not understand "Tully" to refer to the same person as "Cicero," they couldn't be said to know facts about "Tully."

This illustrates the philosopher Gottlob Frege's concept of "sense," the thought conveyed by a word that refers to something in the world. Words that refer to the same thing (the same "reference") can have different senses. The sense is a thought capable of being conveyed and understood by another. It's what we mean when we refer to a thing in the world—or a reference.

Frege explained how the same reference can hold many different senses (which I may also describe as “thoughts”) or "modes of presentation." In language, it's not necessarily what we describe that matters, but how we describe it. Reference and sense can be represented as the difference between "denotation," the literal thing that a word refers to, and "connotation," the idea that the word stands for, respectively. A word that "denotes" a single thing can "connotate" many different ideas standing for that thing in different ways.

Take yourself, dear reader. I can refer to you by your profession (ex: the Lawyer), by your alumni status (ex: the UChicago Graduate), or by your relationships (ex: the Boyfriend). Each of these is valid when referring to "you.” And this is not a contradiction—all of these senses of you make sense in the right context. Just because you are a lawyer and a boyfriend doesn't mean that lawyer = boyfriend. Rather, both are possible ways of framing you, my hypothetical reader. They are different “senses” of “you.” 

Now that we understand that a term can convey multiple senses, we can resolve some hot-button national issues. Like, "What is milk?"

What is Milk?

Milk is defined as "an opaque white fluid rich in fat and protein, secreted by female mammals for the nourishment of their young." However, how can we think of products like soy milk, almond milk, hemp milk, and pea milk as “milk” if they do not derive from a mammal? Based on the definition of milk, isn't a term like "almond milk" a contradiction, like a "square circle"?

As I have argued previously, if a phrase conveys a thought, it's not meaningless. When saying "almond milk," we would understand "milk" only as a white beverage with a slightly sweet and creamy taste, regardless of whether it came from an animal—dictionary definition be damned.

The "milk" characterization of “almond milk” presents a sense of almond milk's qualities as a drink rather than its origins. In the context of drinking, if it looks like milk and tastes like milk, then it is "milk." 

Meanwhile, a term like "mother's milk" expresses a sense of how the milk is produced. Depending on the context, we can use "mik" to describe either a sweet, creamy beverage or a liquid sourced from a mammary gland.

There can't be an actual contradiction in language—hence the meaning paradox. If you are referring to something with mutually incompatible properties (ex: “colorness green ideas”), but still convey a thought, then there isn't a nonsensical contradiction so long as a listener can make sense of what was said. 

However, there are ways that almond milk is relevantly not milk. We wouldn't treat almond milk the same as cows' milk if we had a severe nut allergy, for example. There, we would care more about the milk's origins than its taste and appearance. "Milk" would not refer to "almond milk" in this context. 

The looks-like, tastes-like test doesn't apply when the milk's source is relevant. And what if someone had asked you for milk to drink, but the only milk you had was well past its sell-by date? This sour milk is technically still "milk"—the dictionary doesn't require that milk be unexpired. But serving your friend old milk would miss the "sense" of the word "milk" he had meant. 

So sometimes, almond milk is just milk. And at other times, almond milk is not milk. Sometimes, milk might not even mean milk. Whether we would classify something as milk or not isn't dependent on milk’s innate features or formal definition, but on how "milk" is used and understood in context.

What is Water?

We can apply the above understanding of using "milk" to philosopher Hilary Putnam's famous "Twin Earth" thought experiment. Putnam asks us to imagine a planet identical to Earth in all respects except for one relevant discrepancy. On this "Twin Earth," the liquid that bears all the physical qualities that we associate with water on Earth (a flavorless, clear liquid) is actually the chemical formula XYZ rather than H20. Although water on twin Earths superficially appears and tastes the same as water on our Earth, it is an entirely different chemical compound that shares the same external features of H20 by coincidence.

Therefore, when someone on Twin Earth refers to "water," are they referring to the same thing we are when we use "water" to refer to H20? Or does the meaning of “water” depend on how it’s used?

Despite XYZ's chemical composition, "water" can stand for XYZ on Twin Earth—depending on the situation. When we refer to a liquid, we can refer to it as a drink, or a liquid with certain chemical properties.

If XYZ looks like water, tastes like water, and has the same consumption effects as water, then what does its chemical makeup matter when we need a drink? If XYZ satisfies this "sense" of the word "water" that the speaker conveyed, then XYZ would be water. For example, if you asked me for water and I filled a glass with water from my toilet, I would have the wrong sense of water that you conveyed, even though it was still H20.

If we were on Twin Earth and said to you, "Please give me some water," you wouldn't necessarily understand that statement as requesting that you return to Earth to give me H20. We may both grasp that I had referred to XYZ as "water" since, in that context, we may understand "water" as a transparent, tasteless fluid for consumption rather than something with a specific chemical composition.

However, in another context, like when working in a chemical laboratory (where we would care about molecular composition) we would not characterize XYZ as water if we had meant "water" to stand for H20.

"Water" can mean one thing in one sense (that includes H20 and XYZ) and yet another thing in another sense (that includes H20 only). Still, there is no contradiction here, as these different senses do convey definite thoughts. The thought becomes definite only in context. 

This isn't a deconstructionist account of meaning, where words mean whatever the speaker says they mean. Words do have a fixed meaning. But that meaning is just as dependent on context as they are on linguistic rules.

This understanding takes what is known as the "encyclopedic" view of meaning in cognitive linguistics, as opposed to a "dictionary" view of language that has been adopted by many analytic philosophers.

Dictionary and Encyclopedic Views of Semantics

Intuitively, words have a definite meaning. Whether or not we are using a word correctly would depend on whether its use is consistent with its ascribed dictionary definition. This is the “dictionary” view of meaning, as used in cognitive linguistics, where lexical symbols are ascribed a single, fixed meaning that sets the standard of proper usage.¹

The “dictionary view” of meaning has been adopted in analytic philosophy, where words are understood to stand for a single representation at all times. Like how the number “2” at all times stands for the natural number between “1” and “3,” words similarly stand for a single representation that fits into a coherent model of language.

However, the dictionary definition of milk can't explain how we refer to "almond milk" as "milk" since it fails to satisfy one of the key properties of milk by not being from an animal. “Almond milk,” is an illogical term that the logic of the dictionary can’t explain.

A dictionary may adapt to how language is used and include these more expansive meanings of “milk” in its definition (like if we were to also include a definition of milk as only a "white, slightly creamy beverage"). But of course, the "sense" contained by the word "milk" is independent of what is published in a dictionary. “Sense” depends on the word's use, and can include "almond milk" if it is successfully used to convey that thought. 

We can’t understand “almond milk” with the dictionary view of meaning, but we can with the "encyclopedic" view. To illustrate this concept, Vyvyan Evans's textbook on cognitive linguistics asks us to think of the word “bachelor.”

While the definition of "bachelor" is a man who has never been married, we wouldn't use the term "bachelor" to describe a priest or a gay man in a committed civil union. 

We may even use "bachelor" to refer to a promiscuous married male, using the word to describe a sex-focused man. The "sense" of a bachelor isn't being "unmarried" but the behavioral characteristics we associate with the quintessential bachelor. So, we can successfully convey a meaningful thought when we refer to a married man as a bachelor. Meaning is truly fluid.

This approach integrates the study of what a word means with the study of how that word is used—or the fields of semantics and pragmatics, respectively. There is no principled distinction between semantics and pragmatics for many cognitive linguists who adopt an "encyclopedic" view of meaning. What a word means is connected with how it's used.

You cannot properly understand English conversations using only the English dictionary. This is why lexicographers examine how a word is used to give it its definition. Once we understand that a term can be used and applied in different senses, we can better understand one another in what I'll colloquially refer to as the "trans-debate."

Social Sex

Anyone living in this day and age is likely familiar with the "trans-debate." I won't rehash it here, but it can be summarized by some taking the view that those who I'll call the "trans-advocates" argue that society should only recognize trans people as the sex they identify as, whereas "trans-skeptics" argue that there are real biological differences between men and women that we should continue to regard. 

The trans-debates appear as a shocking example of how two separate swaths of society, who both have their reasonable proponents and arguments, can manage to speak completely past one another.

Despite our many methods of communication, the ease of global discourse, and our shared values of scientific evidence and the general welfare (the other side has it more than their opponents are willing to admit), the “trans-debate” has been incredibly fruitless. Proponents on both sides are convinced that the other advocates for blatant human rights violations, while failing to acknowledge that the other side still has its share of reasonable arguments that should be given at least some weight. 

You might have expected that our ability to speak so easily with one another would have allowed us to arrive at the same result through reason. Discourse shouldn’t be this hard. I hope someone does a retrospective of this national conversation to assess what went wrong and what problems it has revealed. But if we can't even agree on basic definitions, like what "woman" means, productive discourse will remain elusive. 

I used the "milk" example throughout this piece to illustrate how the literal definition is not how we use the terms "man" and "woman."

Given this substack's theme of addressing overly-debated topics (morality) with what appear to be straightforward answers (contractualism), this article will hope to settle the principles of how this debate should take place. 

How we use the word "woman" can mean different things in different contexts. If I am trying to refer to someone who is wearing a hat and possesses feminine features, I may point out that person and declare, "That woman over there wearing a hat." In that sense, a woman means someone with certain external features representing our general idea of a “woman” (longer hair, softer facial features, smaller frame, etc.).

It wouldn't matter if that person were actually a male actor only dressed as a woman or what type of chromosomes that person had, as referring to someone as a woman successfully conveyed the thought of “that woman.” It’s meaningful, so what is this meaning? These superficial physical qualities of women make up what I'll refer to as the "social sense" of the concept of "woman." Under this concept, if it looks like a “woman,” it is a woman.

Meanwhile, biology would be relevant in other contexts, like when speaking to a healthcare practitioner. “Woman” would then mean a certain set of reproductive and hormonal systems.

Of course, “women” can be meant both in the social and biological sense. But when there is a mismatch between these two sense—like with trans-women, then we need to get specific about which sense of woman we’re using.

There are many other senses of the term "woman," such as when a man refers to another man as "being a woman" if the man is exhibiting feminine behavior. This is using the metaphorical sense of “woman” to attribute feminine qualities to someone unbefitting for a man. And since it conveys a thought, it is a meaningful use of the word. But for our purposes, we can stick to only using the "social sense" and the "biological sense" of women.

The term "woman" is underspecified. When we say “woman,” do we mean its social sense? Biological sense? Both? Since it is capable of standing for a variety of ideas, we can't be sure what is meant exactly when spoken.

However, it is given a specified meaning based on context, which may or may not include trans-women. The primary issue in the trans-debate is specifying which senses of the word "woman" we should use in what situations. Sometimes, the social sense of sex will be relevant, and other times, it’s the biological sense—which would exclude trans-woman. 

Again, this is not deconstruction, where the terms "man" and "woman" mean nothing and are therefore capable of meaning anything. Rather, the term "woman" contains a set of different ideas. Context determines which ideas are meant when we use "woman." Sex is a "social" construct, not a personal one. There is a definition of "women," but just because it means different things in different circumstances doesn't make it meaningless, or capable of taking on any meaning ascribed. Objective meaning still exists. 

In a public setting, someone who appears and acts as a woman can be correctly described as a woman; they would fit into one of the concepts of a woman. Most of the time, we don't care about biological differences and just use the social sense—how someone appears to the world—to refer to a woman in public. Yet to a gynecologist or a man seeking a romantic relationship that will lead to naturally born offspring, a trans-woman would not be a woman. These circumstances adopt a certain definition of woman that biological men fail to satisfy.

The "trans-advocates" focus on the social sense of sex, whereas the "trans-skeptics" focus on the biological sense of sex. And no side is absolutely wrong (meaning wrong in all circumstances), but relatively wrong (wrong only in some circumstances). Almond milk can be milk in one sense and not milk in another. A person can be described as a man in one sense and a woman in another sense. There are no contradictions since both words can be spoken and understood in different senses. The debate on trans issues can be more constructive if we accept this principle, adopting both sides' stances.  

So, what sense of the word "woman" should we use? It depends on the situation and what "woman-like" qualities are relevant. Suppose a non-athletic social club that was female-only excluded a trans-woman from being a member. In that case, we may think that the social sense of "woman" would be relevant, rather than the biological sense, making trans-exclusion discriminatory. The hormones or bone structure of a trans-woman seems irrelevant to her qualifications to join an all-female club when the trans-female looks like and behaves like a woman. Trans-skeptics should accept this point, whereas trans-advocates should similarly recognize that "biological sex" does become relevant in certain areas, like athletics. Although requiring full inclusion of trans-women could be acceptable for a social club, the physical differences between men and women become relevant considerations in sports. 

Biological Sex is (Mostly) Irrelevant

Where do we go from here? Well, for the most part, trans-advocates are correct—biological sex is irrelevant, at least from a political standpoint. Legal distinctions on the basis of sex are deemed unconstitutional under the 14th Amendment unless the government proves that such a classification would further an "important government interest by means that are substantially related to that interest." This is a demanding standard that extremely few sex-based classifications have survived—what is known as intermediate scrutiny. Yet athletics and bathrooms have met this standard.

We don't see separate bathrooms and sports teams as a form of "separate but equal" but as justified separation based on biological differences between the sexes. The trans-debate at least has been largely confined to these domains, as is appropriate. So these areas can accommodate legitimate discussions around gender. We should now ask how should our existing legal rules that account for biological sex differences be applied to trans-people. Once we answer this, the trans-debate will be resolved. 

Sex Relevant Questions

The trans-advocates should accept that there are biology-relevant situations where the biological sense is at least somewhat relevant. Some ground should be given on this point that trans-women in female sports is less clear than trans-women in female social clubs.

By defining our terms and recognizing the different senses of "woman," we would be in a better position for discussion. We can agree on what scenarios would be best to classify "women" in its biological sense instead of its social sense. And if we use the biological sense, we would give at least some weight to biological-based reasons (opportunity, safety, etc.) and establish rules that balance these legitimate concerns with those of equality, fair treatment, and inclusivity.

None of these concerns are absolute. I assume they wouldn't justify either complete inclusions or exclusions of trans-women from the definition of "woman" in contexts where biology is relevant, such as sports. The result would likely be carefully crafted rules that address both sets of concerns as well as possible.

At the very least, the rationales provided by both sides should be understood and given weight. The chances are slim that one side or the other is entirely right. It’s far more likely that there are reasonable arguments going in both directions and just need to be reconciled.

Conclusion

Words do not have an absolute meaning. They have meaning only in context—they’re versatile containers capable of unbounded expression. How the word is actually used determines the exact idea that’s expressed.

The word “woman” is no different. It can stand for multiple ideas in multiple settings. We now just need to examine what we mean by “woman” in each of these settings. Once we match the right sense of woman with each of these settings, we can resolve our broken dialogue.

1. This section is informed by Chapter 15 of Vyvyan Evans’s Cognitive Linguistics: A Complete Guide.