Symbols and Language

Values, norms and sanctions are the machinery of conduct. None of it could operate for a second without the medium that carries it: a system in which things stand for other things.

That is what a symbol is, and it is the most distinctively human thing about human beings.

Consider a rectangle of dyed cloth, roughly a metre by a metre and a half. Cotton or polyester. Manufacturing cost: very little. Its physical properties are unremarkable — it will not keep you warm and it tears.

Now: people have died for that piece of cloth. Not metaphorically. Soldiers have been killed retrieving one from a battlefield. Countries have laws about how it must be folded, at what height it may fly, when it must be lowered, and what may not be printed on it. Burning one can end a career, start a riot, or constitute a criminal offence. People cry when they see it raised while a particular tune is played. Athletes wrap themselves in one at the moment of their greatest happiness.

Nothing in the cloth explains any of this. Cut it into strips and it is rags. The identical cloth in a different arrangement of colours means nothing at all, or means something to an entirely different set of people, who would fight the first set.

And here is the part that should stop you: you can feel it. If someone described mistreating your flag, something would move in you — and you know, intellectually, that it is a piece of cloth.

The idea, in plain words

Four properties of symbols

Arbitrariness. The word "dog" has no doggish quality; chien, kutta and inu do it equally well. White signifies mourning in parts of Asia and celebration in Europe. A thumbs-up is approval in many places and an insult in others. Nodding means yes in most places and no in parts of Greece and Bulgaria. Because the link is arbitrary, it must be learned — and can be got wrong.

They allow reference to the absent. This is the property that makes culture cumulative (2.1.1). With symbols you can discuss last year, next year, a person who died before you were born, a country you have never seen, a god, a number, or something that did not happen. No other capacity has changed the world more.

They carry emotion, not just information. The flag does not inform you; it moves you. This is because symbols are learned in emotionally charged settings — the anthem at school assembly, the funeral, the wedding, the crowd. Durkheim's account of collective effervescence (4.2) is precisely about this: the heightened feeling of assembled groups becomes attached to the object present at the time, which then carries the group's emotion afterwards. A totem, a flag, a cross, a name.

They can be fought over. Because symbols carry emotion and identity, controlling their meaning is a political prize. Whose statue stands in the square, what a street is named, which script a language is written in, whether a garment is oppression or piety, what a colour means this year. Symbolic conflicts are often dismissed as trivial — "it's just a name" — by people who have not noticed that symbols are how groups are constituted at all (9.5).

Language: three sociological things worth knowing

1. Language is the main channel of socialisation

You did not learn your society's values from lectures. You learned them embedded in words. Every language hands you a set of ready-made categories — of kinship, of respect, of politeness, of what counts as a proper question — and you absorb the categories along with the vocabulary.

Consider kinship terms. English has one word, "uncle," for four quite different relationships. Hindi distinguishes chacha (father's younger brother), tau (father's elder brother), mama (mother's brother), and phupha (father's sister's husband) — and these are not synonyms, because in a joint-family system the relationships carry different obligations, authority and inheritance implications. The vocabulary is a map of the social structure, and a child learning to speak is learning the structure (Karve on regional kinship systems, 6.2.8; 9.1).

The way you speak is one of the most efficient status signals in existence, and it operates in fractions of a second.

Accent, vocabulary, grammar, code-switching, which language you use for which purpose, whether you can produce the register expected in an interview — all of these are read by listeners as evidence about class, region, caste, education and origin. Bourdieu (5.10, 2.1.13) called this linguistic capital, a component of cultural capital: the linguistic competence rewarded by schools and employers, unequally distributed, and mistaken by everyone involved for intelligence.

The consequences are large and measurable. Children taught in a language that is not their home language start at a disadvantage that compounds for years. In postcolonial societies the language of elite education is frequently the coloniser's language, which converts a linguistic inheritance into a class filter (9.2, 6.7). Ngũgĩ wa Thiong'o's argument in Decolonising the Mind — that language is the carrier of a people's collective memory and that colonial language policy was a strategy of domination — belongs in the same conversation.

Note also the vocabulary of respect built into many languages: Hindi's tu/tum/aap, French tu/vous, Japanese honorific registers, Tamil and Telugu address forms. In each, every utterance requires you to encode a relationship. You cannot speak without positioning yourself, which means the hierarchy is refreshed in every sentence.

3. Naming is power

Who gets to name a thing exercises real authority over it. Compare: freedom fighter and terrorist; undocumented migrant and illegal alien; scheduled caste, Harijan and Dalit — where Dalit, "broken" or "ground down," was deliberately chosen by the community itself in place of a name given by others, and that choice was the point (6.2.7).

Foucault's discourse (5.9, 1.5.2) is the systematic version of this: the categories available in a setting determine what can be thought and claimed there. If a form has no box for your situation, you do not exist administratively.

Does the language you speak shape the thoughts you can have?

The claim is associated with Edward Sapir and Benjamin Lee Whorf, and it comes in two strengths.

Strong version (linguistic determinism): language determines thought; what your language cannot express, you cannot think. This is now generally rejected. The evidence against it is straightforward: people routinely think things their language lacks a word for, coin new words, and learn concepts from other languages. If the strong version were true, translation and cross-cultural learning would be impossible, and they plainly are not. Whorf's most famous illustrations — including claims about Hopi conceptions of time and the often-repeated "Eskimo words for snow" — have been substantially criticised, and the snow example in particular became a popular myth through repeated exaggeration.

Weak version (linguistic relativity): language influences habitual attention and categorisation. This has real experimental support. Studies find measurable effects of a language's colour vocabulary on speed of colour discrimination; of grammatical gender on how speakers describe objects; of spatial systems (languages using absolute compass directions rather than left/right) on spatial memory; and of how a language encodes number on certain counting tasks.

The honest summary: language does not build walls around thought, but it does lay paths — it makes some distinctions habitual and cheap, and others effortful. That is a modest claim with good evidence, and it is quite enough for sociology, because habitual attention is exactly what culture consists of.

Sociologically the more important point is the one Bourdieu and Foucault make: whatever language does to cognition, it certainly distributes authority, and the available categories determine what can be officially said.

Go deeper

Mead and Blumer: symbols are what make society possible

The micro-sociological tradition (5.3) makes the strongest possible claim about symbols, and it is worth stating clearly.

George Herbert Mead argued that what distinguishes human interaction is the significant symbol — a gesture that means the same thing to the one making it and the one receiving it. Because you and I share the meaning, I can anticipate your response to my gesture before I make it, and adjust. That single capacity makes possible role-taking, planning, cooperation, and eventually the self (2.3): I can regard myself as others regard me only because I share their symbols.

Herbert Blumer compressed this into the three premises of symbolic interactionism: people act towards things on the basis of the meanings those things have for them; those meanings arise out of social interaction; and they are handled and modified through interpretation. In this tradition, society is not a structure that contains symbols — it is a continuous process of symbolic interaction, and the structures are its sediment.

Symbols and the sacred

Durkheim's Elementary Forms of Religious Life (4.2) makes the boldest claim about symbols in the discipline: that religion's real object is society itself, and that the totem — the emblem of the clan — is society made visible so that it can be worshipped. The sacred is not a property of the object but a status conferred on it by collective feeling.

You do not have to accept his theory of religion to see how much it explains. The flag, the anthem, the constitution, the founder's photograph, the two minutes' silence, the martyrs' memorial: modern states have a full apparatus of sacred symbols, and the phrase for it — civil religion — was coined for exactly this (9.3, 9.5). Notice that the emotions involved are not weaker than religious ones.

Symbolic boundaries and symbolic violence

Two concepts you will need repeatedly.

Symbolic boundaries (Michèle Lamont, 5.12) are the distinctions people draw to sort the world into us and them: what kind of person eats that, dresses like that, speaks like that, finds that funny. These boundaries do practical work — they govern who is invited, trusted, hired, married. Studying how ordinary people draw them is one of the most productive areas of contemporary cultural sociology.

Symbolic violence (Bourdieu, 5.10) is the imposition of categories that the dominated accept, and which devalue them in their own eyes. A child taught that their home language is "broken" or "rustic" learns to be ashamed of their parents' speech. No force is applied. The categories do the work, and the person applies them to themselves — which is why Bourdieu called it violence.

The material carries the symbolic

A last connection back to 2.1.1. Symbols are not confined to the non-material layer; they are lodged in objects, spaces and bodies.

Clothing is the obvious case: uniforms, veils, turbans, wedding garments, the specific status vocabulary of a sari or a suit. So is space: which entrance you use, which room you may enter, where the seating puts you (recall the village meeting in 1.4.4 — the floor and the chairs were a symbolic system). So is the body: posture, tattoos, hairstyle, a sacred thread, a caste mark, the way one carries oneself, which Bourdieu called the social made body.

This is why the material and non-material cannot be separated in practice, and why sociologists read rooms, clothes and bodies as texts.

"Symbols are just conventions, so they don't really matter." Arbitrary and consequential are compatible. Money is arbitrary too (1.3.3).

"Sapir–Whorf proves language determines thought." The strong version is rejected; the weak version — language influences habitual attention — has good support. Say which one you mean, and be sceptical of the snow anecdote.

"Some languages are more logical or advanced." Linguistics finds no basis for ranking languages by expressive capacity. What differs is institutional status — which language the courts, universities and employers use — and that is a political fact, not a linguistic one.

"Symbolic issues are a distraction from real ones." Frequently said of statues, names, flags and dress. Symbols are how groups constitute themselves and how membership is signalled; conflicts over them are conflicts over who belongs.

"If we all just used clearer words, misunderstanding would end." Meaning is not a property of words but of shared interpretive practice. Two people can use identical words and mean different things — which is Blumer's point, and the reason interpretation cannot be engineered away.

Why it matters

It explains the most puzzling human behaviour. Why people die for cloth, riot over a name, feel their throat tighten at a tune, refuse food that would nourish them, or spend a fortune on a wedding (1.3.6). None of it makes sense in a world of objects. All of it makes sense in a world of symbols.

It shows you where power hides in plain speech. Accent, register, naming, address forms and available categories are doing distributive work continuously — in classrooms, courtrooms, clinics and interviews — and almost nobody involved experiences it as power.

And it is the doorway to Part 5's micro-sociology. Mead, Blumer, Goffman, Garfinkel and Sacks all build on a single premise you now hold: that social life is conducted in symbols, and therefore has to be interpreted rather than merely observed.

A symbol is anything a group treats as standing for something else. The link is arbitrary, conventional and learned.

Symbols allow reference to what is absent — the capacity that makes cumulative culture possible — and they carry emotion, which is why they are fought over.

Language is the most elaborate symbol system: it is the main channel of socialisation, it hands you ready-made social categories (kinship terms map social structure), and it marks and reproduces position through accent, register and linguistic capital.

Naming is power: freedom fighter/terrorist, Harijan/Dalit. Foucault's discourse is the systematic version.

The Sapir–Whorf hypothesis: the strong deterministic version is rejected; the weak relativity version — language shapes habitual attention — has experimental support.

Mead's significant symbol makes role-taking, cooperation and the self possible; Blumer's three premises found symbolic interactionism.

Durkheim: the totem is society made visible; modern states run a full apparatus of sacred symbols (civil religion).

Symbolic boundaries (Lamont) sort us from them; symbolic violence (Bourdieu) is domination through categories the dominated accept.

Symbol — anything that stands for something else by shared convention.

Arbitrariness — the absence of any intrinsic connection between a symbol and what it signifies.

Language — a rule-governed symbol system permitting reference to the absent, abstract and hypothetical.

Significant symbol — Mead's term for a gesture meaning the same to sender and receiver, enabling role-taking.

Symbolic interactionism — Blumer's framework: action follows meaning, meaning arises in interaction, and meaning is modified through interpretation.

Linguistic relativity (weak Sapir–Whorf) — the claim that language influences habitual perception and categorisation.

Linguistic determinism (strong Sapir–Whorf) — the rejected claim that language determines the limits of thought.

Linguistic capital — Bourdieu's term for socially rewarded linguistic competence.

Code-switching — moving between languages or registers according to setting and audience.

Discourse — Foucault's system of statements and categories governing what can be said in a setting.

Totem — Durkheim's emblem of the group, through which the group venerates itself.

Collective effervescence — the heightened emotion of assembled groups, which attaches itself to symbols.

Civil religion — the sacred symbols, rituals and narratives of a nation or state.

Symbolic boundary — a distinction used to sort people into categories of us and them (Lamont).

Symbolic violence — domination exercised through categories the dominated have internalised (Bourdieu).

One — the arbitrariness test. List five symbols you respond to emotionally (a flag, a colour, a song, a logo, a garment). For each, find a place or period where the same thing means something else, or nothing.

Two — the kinship map. Write out every kinship term in your first language. Where it makes distinctions English does not, ask what different obligation attaches to each. You are reading a social structure out of a vocabulary.

Three — register switch. Notice how differently you speak in three settings: with a grandparent, with close friends, in an interview. Write one sentence in each. What is being signalled, and what would it cost you to get it wrong?

Four — find a symbolic boundary. Identify one taste, habit or preference that, in your world, marks someone as "not our kind." Then ask what practical consequences that boundary has — who is invited, trusted, hired, married.