Words as Tools: Meaning, Mediation, and Activity

TheThe companion page on meaning and ontology treats meaning as a three-way relation between a sign, an object, and an interpretant, shaped by context and by the practices in which a word is used. That account explains a good deal, but it still tends to picture meaning as something that happens in an interpretation: a word is produced, and someone works out what it refers to. This page takes a different route. It looks at meaning as something produced through activity, where words are tools that people use to act on objects, and where that use changes the people, the environment, and the words themselves.

1. From Interpretation to Activity FoundationalKnowledge that endures for decades — core principles

InIn Lev Vygotsky's account of mediated action, a tool stands between a person and the object of their activity. Tools include physical implements such as a hammer or a map, and they include signs: a knotted handkerchief used as a reminder, a number system, a written note. Vygotsky argued that people use these tools not only to act on the world around them but to organise their own thinking, and that this is how higher psychological processes develop[1].words are tools too. they shape thought.

Later activity-theoretical work, most influentially Yrjö Engeström's, widens this into a whole system. An activity includes a subject, a tool, and an object, and it is also shaped by rules, a community, and a division of labour among the people involved[2].

Text description of the diagram

A triangle with Subject at the top, Tools at the lower left, and Object at the lower right. Below it, a base row joins Rules (lower left), Division of labour (centre), and Community (lower right). The diagram is schematic and omits some of the connections in Engeström's full model.

The important feature of mediation is that its effects run in both directions. People transform their environment through tools, but the tool-mediated activity also transforms the people, the practices, and the possibilities for action that come afterwards. A tool is not a neutral pass-through. Once used, it changes what the user can see, do, and expect.

2. Words as Cultural Tools Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

LanguageLanguage is itself a cultural tool, and the same mediating relationship applies to words. A speaker and their community use a word, adapting it as they go. The word frames, enables, and constrains the way they attend to an object or take part in an activity. The activity that results can then change the object, the participants, and the word.

The common description of words as "tools of communication" is useful but a little narrow. Words do more than carry information that already exists in one person's head. They help people:

• direct attention;
• distinguish one thing from another;
• coordinate collective action;
• establish roles and obligations;
• remember and externalise experience;
• frame problems;
• imagine possibilities;
• classify people and objects;
• negotiate what counts as relevant or real.

Introducing a term does not simply attach a label to an object that was already self-evident. It gives people a way to notice patterns, compare experiences, assign responsibility, carry out research, and organise responses. Technical debt is a useful example. Ward Cunningham introduced the metaphor in 1992 to describe the cost of shortcuts in code[3]. The term then let developers notice a recurring pattern, argue about when to pay the cost down, and plan work around it. The phenomenon was not invented by the metaphor, but the activities the metaphor made possible changed what developers saw in their codebases and how they organised their work.words frame reality, not just describe it

A word is therefore a semiotic or cultural tool. It does not only describe activity from outside. It can become part of how that activity is organised.

3. A Threefold Transformation Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

LinguisticLinguistic mediation involves at least three transformations happening at once. They are different in kind, and each one can be studied separately.

3.1 Words transform their users

Learning a concept can change what a person is able to perceive, distinguish, remember, and do. Before acquiring a specialist concept, someone may experience several situations as unrelated. Once the concept is available, those situations can become instances of one pattern. The word does not necessarily create the underlying phenomenon, but it changes the person's capacity to engage with it, and that capacity is what the activity then draws on.

3.2 Words transform the social and material environment

Words become embedded in institutional rules, professional practices, software interfaces, database categories, legislation, educational curricula, diagnostic procedures, and organisational roles. At that point language has material consequences. A category in an information system can decide which services a person may access. A legal definition can create rights and obligations. A classification used by an algorithm can affect what becomes visible, actionable, or measurable. Meaning, in this sense, is part of intervention as well as representation.

3.3 Use transforms words

A word's history of use accumulates around it. New communities adopt it, new technologies make it operational, and new situations expose limits in earlier senses. The word agent is a clear example:

• in philosophy, it may mean an entity capable of intentional action;
• in sociology, it appears in discussions of agency and structure;
• in software engineering, it may mean a process that acts on behalf of a user;
• in contemporary AI, it may refer to a system combining models, tools, memory, planning, and interaction with an environment.

These senses are neither wholly unrelated nor simple instances of one timeless essence. They form an evolving family of uses, each shaped by the activity in which it developed.cf. Wittgenstein's family resemblance

4. Meaning as Temporary Stabilisation FoundationalKnowledge that endures for decades — core principles

PuttingPutting these pieces together gives a more careful formulation: meaning is a temporarily stabilised pattern of use, interpretation, and consequence within a historically developing activity system.

The word temporarily carries much of the weight. If meaning were entirely unstable, communication would fail, because no one could rely on a term meaning the same thing from one sentence to the next. If meaning were entirely fixed, conceptual change, metaphor, learning, and disciplinary development would be impossible. Stability is real and it is partial, and it lasts only as long as the activities that sustain it.fixed definitions are a useful fiction, not reality

A community stabilises meaning through repeated usage, shared practices, definitions, examples and counterexamples, institutional standards, teaching, correction, technical implementation, and the relations a concept keeps with its neighbours. Each new use can reproduce the established meaning, modify it, or generate a contradiction that eventually forces the concept to change. The stabilisation is an ongoing achievement rather than a property the word carries in storage.

5. The Tool Carries a History Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

AA word is never a neutral instrument that each speaker picks up afresh. Cultural tools carry traces of earlier activity. A word may contain earlier classifications, disciplinary assumptions, institutional values, metaphors that are no longer obvious, histories of conflict, expectations about legitimate use, and distinctions inherited from earlier technologies.

Take intelligence. Someone who uses the word inherits overlapping histories from psychology, education, philosophy, military analysis, computing, and ordinary language. The meaning of the word in a particular exchange is shaped by its present use, but that use happens against this accumulated history. The tool also makes some actions easier and others harder. Calling a system "intelligent" invites certain comparisons and questions. Calling it "automated computation" invites others. Neither expression is a transparent window onto the object it describes.the word shapes the question you can ask

6. Contradictions Drive Semantic Development Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

AA central idea in Engeström's later work is that contradictions drive development. Tensions among an activity's elements, including its tools, rules, community, and division of labour, can push the activity itself to change[4]. The same idea applies to meaning.

Suppose a community uses chatbot to mean a scripted question-and-answer interface. It then encounters systems that generate novel language, search external sources, execute tools, maintain state, write code, and coordinate specialised components. A contradiction appears between the inherited concept and the emerging object. The community might respond in several ways:

• stretching the old term to cover the new systems;
• distinguishing chatbot from assistant or agent;
• creating new technical subcategories;
• disputing whether the new system belongs to the old category at all;
• redesigning the system in response to the terminology.

In each case, the dispute is not merely a search for the correct dictionary entry. It is part of the collective activity through which the conceptual tool is being revised. Arguments about terminology can therefore be read as evidence that a community is developing its tools, rather than as a failure to agree on what the words "really" mean.cf. the limits of fixed definitions

7. Implications for Ontology Applied / MethodologicalKnowledge with a 5–10 year half-life — stable practice

ThisThis view changes how an ontology should be understood. A formal ontology is not only a static description of a domain. It is also a mediating artefact within an activity system. People construct it, negotiate it, implement it, use it to classify cases, discover its inadequacies, and revise it. The companion page's account of formal ontologiesin computing describes the artefact; this page adds the cycle in which the artefact is used and changed.

Text description of the diagram

Five boxes, top to bottom: domain practice; conceptual distinctions; formal ontology; software and institutional action; changes in practice and domain. Downward arrows join them in sequence. A dashed return arrow runs from the bottom box back up the right side to the formal ontology box and is labelled "ontology revision".

An ontology therefore has a double role. It represents a domain, and it intervenes in that domain by shaping what its users and systems can record, infer, and act on. That is the feedback loop: categories are built to organise experience, and experience then comes to be organised partly through those categories.cf. fixed definitions vs living practice

The intervention is strongest when an ontology is built into software. Once a category becomes a required database field, a menu option, or a rule that an automated system applies, it stops being merely descriptive. It begins to structure the activity around it. This is why the questions the companion page poses about scope, purpose, provenance, and revision matter for implemented systems in particular: the consequences are no longer confined to the document that describes them.

8. What the Argument Supports FoundationalKnowledge that endures for decades — core principles

TakenTaken together, the sections above support six claims about meaning:

• Meaning is performative. Words help produce actions and consequences, not only descriptions.
• Meaning is developmental. Use changes both the users and the conceptual tools they work with.
• Meaning is materially embedded. Classifications become part of interfaces, institutions, rules, and infrastructure.
• Meaning is collective. Communities stabilise and renegotiate meanings together.
• Meaning is historical. Every conceptual tool carries traces of the activities that shaped it.
• Ontology is active. A formal ontology does not only model a domain. Its implementation can reshape that domain.

Words are cultural tools, and tools do more than carry intentions that already exist. In use they mediate the relationship between people and the objects of their activity. They direct attention, enable distinctions, organise cooperation, and shape which actions seem possible. In doing so they transform their users and the social and material environments those users work in.

The relationship is recursive. Communities also adapt their words through use. New practices stretch inherited meanings. New technologies turn metaphors into technical categories. Conflicts expose hidden assumptions. Formal definitions stabilise distinctions for particular purposes, and only for as long as those purposes last. Meaning is therefore neither an essence contained inside a word nor an unconstrained choice made by an isolated interpreter. It is a historically developing achievement of mediated, collective activity.meaning shifts with practice, not fixed

Related Topics

• Meaning, Ontology, and the Limits of Fixed Definitions — the companion page on sign relations, context, and formal ontologies, which this page extends into activity.
• Frame Analysis for Software — the requirements-disagreement failure mode that frames and activity systems help explain.
• Informatics, Frame Analysis, and Ontologies — the software-side account of formal ontologies that section 7 builds on.

References

1. L. S. Vygotsky, Mind in Society: The Development of Higher Psychological Processes, eds. M. Cole, V. John-Steiner, S. Scribner, and E. Souberman, Harvard University Press, 1978.
2. Y. Engeström, Learning by Expanding: An Activity-Theoretical Approach to Developmental Research, Orienta-Konsultit, Helsinki, 1987.
3. W. Cunningham, "The WyCash Portfolio Management System," OOPSLA '92 Experience Report, 1992. https://doi.org/10.1145/157709.157715
4. Y. Engeström, "Expansive Learning at Work: Toward an Activity Theoretical Reconceptualization," Journal of Education and Work 14(1), 2001, pp. 133–156. https://doi.org/10.1080/13639080020028747