You open a dictionary and all the words are defined by other words. If concepts have the same circularity as words, ultimately none would have meaning (I suppose that's debatable, but I'm assuming it here). Therefore, it seems like there must be some mental firmware or foundational concepts to ground them all in reality and provide meaning. Would such concepts be undefinable even though we might refer to them? For example, try to define 'time' without using other temporal concepts. Are all concepts definable or are some irreducible?

  • 1
    @curiousdannii My question assumes circularity leads to meaninglessness before asking about something like fundamental concepts. The link about fundamental concepts was very helpful. Particularly "semantic primes" looks like what I'm getting at. Thank you.
    – Mac
    Mar 6, 2022 at 17:43
  • You point out only troubled definitions in the dictionary. Most of the words have legitimate definitions with detailed explanations.The words without detailed information typically use SYNONYMS as the definition. Take the term HUMAN BEINGS & you will find synonyms only. THAT will lead to circular reasoning if you argue word definitions using the dictionary. You shouldn't argue about definitions using the dictionary in the first place. You would be committing a logical fallacy arguing from the dictionary with many words & you seem to notice this. Many words are defined well though. Some aren't
    – Logikal
    Mar 6, 2022 at 19:07
  • On the topic of mental firmware, image schemas are a related match in cognitive linguistics. I outline their basis to higher concepts in this answer.
    – Michael
    Mar 6, 2022 at 21:35
  • This is perhaps the oldest question in philosophy. What is the underlying substance or certainty from which all else can be derived? Perhaps simply existence or "being." Kant revolutionized our approach to this question by proposing and demonstrating that, to simplify greatly, there are a number of fundamental concepts or "categories" without which we cannot think, imagine, or even perceive, time and space being two such categories. The firmware of any rational being. While most philosophers would not accept this today, it is a path out of the relativism and circularity you describe. Mar 6, 2022 at 21:38

3 Answers 3


Analysis of natural language probably belongs to linguistics rather than philosophy. The linked question Since words are defined in terms of other words in dictionaries, leading to infinite loops, does it mean natural languages are meaningless? has typical answers to the general question. Nevertheless the question here is phrased as a problem for directed graphs, and some logical statements can be derived there.

Formally, even if words could be reduced to other words by definition, it is possible that some definitions are circular, and so it is not necessarily true that any words of a language are irreducible.

Even if words could be reduced to other words by definition, and there were words that could not further be reduced, that does not mean those words cannot be defined. Such words could still be "defined without reduction". The relationship "X reduces Y" and "X is used in some definition of Y" should be independent (it seems to me). So irreducible is not necessarily the same as undefinable.

A key point here is that a definition of a word in natural language is typically not unique, so the same word could be defined in two similar ways using very different words. So reducing by means of definition seems inadequate for most purposes. The relationship between concepts and the relationship between the word used for the concepts are not strict enough in natural language to make useful statements.


The comments above are helpful, let me make one addition which may be of interest.

  1. Definability has been well- treated in the literature. In particular, we have a number of constraints on when a certain concept is definable with respect to a background language.

The high level notions (on a particular notion of definability) are these: that the concepts be "explicable" purely in terms of the ground language, and that they add no additional inferential power to the language.

For a brief introduction, see as always : https://plato.stanford.edu/entries/definitions/.

  1. Seen thus, your question is one of asking if the concepts of the base language are definable in terms of a more primitive language. It is easy to give examples in which this is true - merely extend any language twice, and take the given base language to be the second extension. What you really wish to know though, is whether given that there is indeed a truly primitive language- whether we can define in it terms other than itself. If we allow the definability relation to be trivial, this is the case, of course, if the language is indeed truly primitive, there is no other language in which we can express its terms (or should be, I haven't worked this out myself).

Thinking of Feynman’s words on ‘what is energy’, we can try as hard as we like to understand the physical world but however thorough or incomplete our understanding we must eventually put up with the idea that it just is what it is, and so ultimately we have no irreducible basis for anything. Even ’cogito ergo sum’ isn’t irreducible because even though we are logically obliged to agree that this must be true, we can still ask questions about thought and consciousness. So our language doesn’t exist in a vacuum but is based on the idea that things are just as they appear to be, which is not irreducible but it’s all we’ve got.

  • Energy does seem like one of those hard to define concepts. There must come a point where thought ends and reality continues. However, I didn't really understand the rest of the answer about there being nothing irreducible.
    – Mac
    Mar 6, 2022 at 17:47

Not the answer you're looking for? Browse other questions tagged .