Five languages for describing a group, and when to switch

One author counts observable facts, one individual case at a time: this person, this situation, this reaction. Another describes the same processes structurally, in the language of systems, levels and functions, where the individual case nearly disappears. A third writes about ant colonies and neural networks so plainly that the reader suddenly recognises their own sales department in them. All three are describing the same thing: how separate individuals turn into a stable “we”. No single language takes the subject whole, which is why social psychology rests on several different supports instead of one.

That reads like a problem for a university reading list. It is a working problem for anyone trying to understand a market, and it gets sharper when the market is Russia and the person doing the understanding sits in another country with a dashboard open.

A subject that does not fit inside one discipline

The way I teach this material, it is built deliberately around a set of sources with different optics rather than around one textbook.

There are texts sharpened on concrete observable facts and the individual perspective: what happens to one person in one situation. There are their structural counterparts, which describe the same reality in the language of systems, levels and functions instead of separate cases. There are books on small, medium and large groups, because the regularities of a pair, a department and a crowd overlap only partly. There are semi-literary texts about network structures: how communities are wired in biological systems and in human society, where the topology of the network explains the behaviour of the group more precisely than the psychology of its individual members. Separately there is axiology, the system of values, without which the structural books read as bare mechanics with no meaning attached. And two genres on the border: biopolitics, which ties political acts to neurohumoral regulation, and the popular ethology of politics, meaning how the biology of behaviour shows up on the public stage.

None of these texts covers the subject on its own. Facts without structure scatter into anecdotes. Structure without values turns into a cold schematic. A network without an individual history loses the living person. The perspectives have to be stitched together on purpose, and the reason is not that nobody wrote one good book. The reason is that the subject itself has no single language of description.

Two passes over the same material

Behind the variety of sources sits a simple architecture: everything gets covered twice.

The first pass is the theory of origin. Where the individual comes from, how identity grows out of the dyad with the primary object, how attitude, attachment, role, leadership, rebellion and separation appear. The second pass takes the same material and turns it instrumentally: how to apply this right now, looking at a specific group and a specific person inside it.

The instrument of the second pass is the sociogram, an extended version of classical sociometry. The method does not return a list of who chooses whom. It returns a full map: the actors, their energy, the boundaries, the direction of attention and of action, plus a separate map of the meaning contexts of the situation. Without the theory of origin a sociogram degenerates into fortune telling by drawing. With it, the drawing becomes a diagnostic language where every detail reads through a mechanism you can name.

Why a book about ants sits next to clinical psychology

It looks illogical to shelve a text about the network structures of biological communities beside books on human psychology. That text gives exactly what purely clinical sources cannot: a view of the group’s topology from outside.

An ant colony does not debate its values or formulate a strategy, and its structure works anyway. Division of function, feedback, emergent behaviour larger than the sum of the individuals in it.

A human group is wired similarly, at a management meeting, in a family, in a circle of friends. What happens at the level of the whole system often cannot be derived from the intentions of the participants. A department head states the goal of the department, and the stated goal almost never coincides with what the department actually is. Texts about networks hand you a language for that gap long before clinical psychology finds a term for it.

What this looks like from outside Russia

Foreign teams usually arrive with one description language fully developed and the rest missing. The developed one is the analytics language: events, sessions, sources, funnels. Years inside Google Analytics and Google Ads train a team to describe an audience as a stream of individual behaviours, and that training is genuinely good. It is also structurally blind to the layer where a group decides.

In Russia that blindness costs more than at home, for one reason that has nothing to do with psychology. A large share of the conversation about your category happens inside Telegram channels and chats, VK communities and closed professional groups. Recommendations circulate there, objections get formed there, and a purchase is often pre-approved there before anyone reaches your site. None of that traffic reports itself back into an advertising account. Yandex Metrica will give you the fact layer with unusual honesty, including session recordings, and it still shows you individuals arriving one by one, never the room they came from.

The instruments a Western marketing team would normally reach for to fix this were built around platforms that carry little of the Russian discussion. Point standard social listening at this market and it returns a nearly empty map, and the natural conclusion is that there is no conversation to listen to. The conclusion is wrong. The instrument is aimed at the wrong place, which is a measurement failure wearing the costume of a market insight.

The values layer fails quietly in a similar way. A message translated correctly can still land wrong, because tone carries a claim about the relationship between seller and buyer, and that claim is read locally. Directness that signals confidence in one market can read as pressure in another; informality that signals warmth can read as a lack of seriousness. Structural insight without a values map is mechanics with the meaning stripped out, and the symptom is a campaign that is configured well and converts badly for reasons nobody in the account can name. This is the part of customer research that no platform report will do for you.

What the local practice actually requires

The practical consequence for anyone sorting out their own work with people: when one description language stalls, the subject is not exhausted. The optics are.

If a conflict in the funnel will not explain through the personal motives of two people, look at it as a structure of roles independent of the specific persons. If the audience will not break down into neat functions, look at it as a network where the job title matters less than who is actually connected to whom by a flow of information. If the structure is clear and the behaviour still surprises you, the missing layer is values.

For a company selling into Russia from abroad, that translates into a short and unglamorous sequence.

1. Ask the origin question before the size question. What did this audience form around, which shared situation gave it a shared vocabulary. Market volume tells you nothing about how the group is wired.
2. Map the actors, not the segments. Who decides, who blocks, who certifies, and which intermediaries already stand between you and the buyer. In Russian search results aggregators and marketplaces are actors in your sale whether you acknowledge them or not.
3. Read the language people actually use. Wordstat phrasing and community threads give you the words your category is argued about in, which is a values signal as much as a keyword list.
4. Check who answers whom. Inside a thread, the person who gets replied to is the node with the energy. That is a topology fact and it is visible without any paid tooling.
5. Keep both passes running. Origin explains why an argument works. The instrumental map tells you where to place it this quarter. Either one alone produces confident nonsense.

This also removes the temptation to hunt for the single correct theory of personality or the single correct method of working with a group. The language of facts and separate cases, the language of structures, the language of group regularities, the language of networks and the language of values do not contradict each other on substance. They describe the same regularities from different sides. Being able to switch between them is a separate skill, worth developing as deliberately as any specific technique, and it is the skill that decides whether a marketing strategy survives contact with a market its authors did not grow up in.

The map comes before the march

A conversation about a new market starts with a map of sources rather than a finished answer, because the subject demands several approaches from different sides. The sociogram is built the same way: one language speaks about actors, another about contexts, and only the two maps laid over each other produce a picture with depth.

The variety of texts here is not decoration. It is training for the thing itself: holding several perspectives at once without collapsing any of them into the one that is supposedly right.

Frequently asked questions

Sources

• Russian version of this article on belokrylovo.ru: Зачем теме группы и идентичности несколько разных книг