Does language always work like a picture?

A word at the workbench

Nia’s neighbor Greta, the bicycle repairer from the previous lesson, is mending a chair while Nia helps. Asked where the clamp is, Nia replies, “The clamp is beside the lamp.” Later Greta holds two pieces of wood together and calls, “Clamp!” Nia passes it to her.

Both exchanges involve the same tool. In the first, Nia reports where it is; in the second, Greta asks for an action. Knowing which object “clamp” names helps in both, but it does not by itself explain the difference.

Frege’s distinction between sense and reference showed how one object can be presented in different ways. Ludwig Wittgenstein asks us to examine another question: how does language have meaning at all? His early and later work answer it differently, and the workbench lets us follow that change.

How words can picture a situation

In the Tractatus Logico-Philosophicus, Wittgenstein treats a meaningful proposition as a picture of a possible situation. His focus is the statement that can be true or false because it represents how things could stand in the world: its elements fit together to represent that possibility. [1]

“The clamp is beside the lamp” does more than list two objects; it places them in a relation. We understand what would make the statement true whether or not we have looked at the bench, and we can then check it. Perhaps the clamp is actually in a drawer. The statement remains meaningful when it is false.

A logical picture need not look like a photograph. Printed words do not resemble a clamp or a lamp; what matters is a structure through which a possible arrangement can be represented. Say instead that the clamp is under the lamp, and a different situation would make the statement true. This is only a first illustration of picturing, since the early work also demands an analysis into simple objects, which naming two tools does not carry out. [1]

The early question is demanding: what must language and reality have in common for a proposition to represent anything? Describing a situation cannot itself settle whether that situation exists.

Look at what happens next

Return to Greta’s “Clamp!” If Nia answers “True,” she has missed the request; if she brings the tool, she has responded appropriately. Greta is not reporting that a clamp exists or saying where it is.

In the later teaching texts collected as The Blue and Brown Books, Wittgenstein asks how words are explained and used. His imagined builders learn calls for materials along with the actions those calls require, and a name and its object do not exhaust the explanation. [2]

The workshop makes the same point in another setting. Nia must learn when a call asks her to bring something, when it checks a name and when someone is describing the bench. She could picture a clamp during any of these exchanges, and the image alone would leave the difference unexplained.

Meaning within an activity

A language-game lets us examine words together with the activity in which they function. The term can name a deliberately simple case used to investigate language; it does not imply that speaking is trivial or mere pretending. [2]

Greta might teach “clamp” by pointing and asking Nia to repeat it. During a repair, the call asks for the tool; during an inventory, the same word may mark what belongs in an empty space on a list. These are differences in use: the part an expression plays in what people are doing. Understanding the word includes knowing how to respond in each of these circumstances.

In Philosophical Investigations §43, Wittgenstein connects meaning with use for a large class of cases, expressly allowing exceptions. We should not turn that qualified claim into a formula that explains every word in advance. [3]

Use still allows mistakes

Suppose Nia brings a lamp when Greta calls for the clamp. In the workshop as described, that is a mistake, and Greta can correct it by showing the tools and carrying on with the task. The shared practice gives the correction its point.

Now suppose Nia privately decides that “clamp” will mean “lamp.” Her decision does not silently change what Greta asked for. They could agree on new names, but they would have to establish the change together. Meaning in use is not whatever someone happens to associate with a sound, and Wittgenstein’s teaching examples make both learning and the possibility of misunderstanding visible. [2]

A misunderstanding can also concern the activity rather than the object. Nia might correctly point out the clamp while failing to bring it: she knows which tool it is, yet she has not answered the request.

Related without one common feature

Must all these uses share one feature before we can call them language? Wittgenstein questions that demand with the idea of family resemblance: overlapping similarities that need not reduce to a single feature shared by every member. His comparison with games appears in The Blue Book. [2]

Compare a card game played to defeat an opponent, a guessing game in which everyone helps solve a puzzle, and a child throwing and catching a ball alone, over and over. Competition matters in the first and cooperation in the second, while skill and amusement connect other parts of the comparison. We should examine those connections rather than assume in advance that one feature explains every case.

The workshop offers a smaller comparison. Naming, requesting and reporting overlap in vocabulary and in the tools involved, while differing in purpose and in the response each calls for. We can recognize those connections without treating naming and requesting as disguised reports.

An everyday concept can also work without an advance ruling on every borderline case. If the workshop organizes a games afternoon, it can draw a precise boundary for that occasion, but choosing a boundary is different from discovering a hidden definition that every speaker had been using all along. [2]

A change in philosophical attention

The contrast concerns what guides the investigation. The early account asks what shared structure makes representation possible; the later work examines varied uses, and the confusions that come from forcing them into one pattern. Scholars disagree about how sharp a break separates the two projects. [4]

The early text already attends to the use of signs. And in the later book’s preface, Wittgenstein acknowledges serious mistakes in his first book while asking readers to consider the new thoughts against the old. [1] [3]

We still need to ask what would make Nia’s report true, and we also need to understand what Greta is asking Nia to do. The later perspective helps us resist making the first exchange the model for the second. It does not make reporting worthless or turn truth into personal preference.

Try the distinction

A visitor hears Greta call “Clamp!” and says, “That means there is a clamp here. If Nia brings a lamp, that is equally correct, because words mean whatever each person uses them to mean.” Explain the two mistakes. Could the visitor’s sentence ever serve as a request? This exercise is optional.

Read a possible response

The visitor has turned a request into a report. In this situation Greta needs Nia to bring the tool, and merely confirming that one exists does not answer her; Nia could identify the right object and still miss the action being requested. The words have to be understood within what the two people are doing.

The second mistake treats meaning as private. Their established use distinguishes the clamp from the lamp, so a private change of association does not make the wrong tool an equally good response, though they could teach or agree on a new use. As for the visitor’s own sentence, in suitable circumstances “There is a clamp here” could itself prompt someone to fetch one. Its grammatical appearance alone would not settle its role. We would need evidence about the exchange, rather than assuming either that every sentence reports a fact or that any interpretation is acceptable.

Carry this forward

When words puzzle you, ask what people are doing with them and what would count as understanding them. Describe experience before explaining it away turns from the uses of language to the careful description of experience.

Sources and limits

[1] Ludwig Wittgenstein, selections from Tractatus Logico-Philosophicus, translated by C. K. Ogden (1922), in A. P. Martinich and David Sosa, eds., Analytic Philosophy: An Anthology, second edition, Wiley-Blackwell, 2012, for the picture account. Read the complete anthology pages 118–125. This is an edited selection, with omissions between numbered remarks, not a continuous reading of the original work. The lesson draws particularly on these remarks:

• Picturing: 2.1–2.225.
• The use of signs: 3.326–3.327.
• Propositions and clarification: the included remarks from 4–4.112.

The decimal numbers identify remarks, not pages. The catalogue link identifies the anthology; it does not supply the text. The workshop illustrates one aspect of the account without identifying ordinary tools with its simple objects.

[2] Wittgenstein, selections from The Blue and Brown Books, in the same anthology, pages 567–569 and 575–578, read completely, for explanation and use, language-games and family resemblance:

• Explanation and use: the opening discussion.
• Language-games and family likenesses: pages 576–577.
• The builders: pages 577–578, where the example begins.

The anthology credits the 1958 volume. Its original pagination was not independently checked; these locators refer to the anthology. Passages continuing beyond the selected pages are not used as completed arguments.

[3] Wittgenstein, Philosophical Investigations, revised fourth edition, Wiley-Blackwell, 2009, translated by G. E. M. Anscombe, P. M. S. Hacker and Joachim Schulte, revised by Hacker and Schulte, for meaning as use and the comparison with games. Read the complete English preface and §§1–23, 38–43 and 65–71, including the embedded boxed remarks in those ranges. Section 43 qualifies meaning as use; §§65–71 develop the comparison with games and the question of boundaries. Section numbers are more useful here than PDF page numbers. No German or earlier-translation comparison was made.

[4] Anat Biletzki and Anat Matar, “Ludwig Wittgenstein”, Stanford Encyclopedia of Philosophy, substantive revision 17 September 2026, for the early account, interpretive disagreement and the later methods. Read the complete main text of §§2.1, 2.4 and 3.1–3.4. Works discussed there were not thereby read directly.

Sources consulted 18 September 2026. Nia, Greta, the repair-workshop exchanges and the optional exercise are original teaching material; the builders and the family comparison come from Wittgenstein. This bounded reading does not cover either complete book, the private-language debate or every interpretation of the early work. No independent scholarly review or learner testing is claimed.