An infinite regress in linguistic definition?

You are correct of course but this ignores the fact that "words" like "numbers" are generally used in real life situations of mutual context which exclude the need for "definition". Wittgenstein pointed out the futility of much of "philosophy" because such context is often lacking...hence his dictum "meaning is use" (i.e. use in context)

EP, I like your name. Welcome to a2k.

I think you are right. This is why a non-English speaker can't learn English directly from a dictionary - they have to slide from their infinite regression into ours.

Now we have the problem of how does one first learn words if one needs to know words in the first place? Obviously, our first words we learn the definition of, not by words, but by sensations - sight, smell, sound.

Exactly, language is created to express different sensations and ideas so after that everything builds up. For example you catergorise foods with a certain taste as being sour than you realise these foods are different for example a Lemon so you learn that word etc.

so the defintion of words are ultimately based in sense perception.

so was the first evey language descriptive?

or is that a stupid question?

Only for some words. There are other words whose definitions derive as much from the structural or syntactical needs of the linguistic system itself as from sense experience--things like articles, some interjections, etc.

The term "sense data" cannot be taken as a priori because "data" depend on the activities and needs of particular observers in context. Better to think of language as vocal extensions of situational relationships which later become internalized and representational of those relationships themselves.

Thus "thought" can be considered to be "sub vocal speech". As shapeless implied above, the syntax of socially transmitted language "shapes" our subsequent "thinking" as reflected in concepts like "causality" which is directly related to the Subject-Verb rule for minimal sentences. The philosophical implications of this are profound in as much that "causality" may simply be an artifact of "language" which serves the particular "needs" of our species to "predict and control".

Did not Hume show the causality was just an habitual thing that we humans use to explain the world?

Precisely. The term "explanation" usually implies "causality" which Hume was unable to account for in philosophical terms. Kant argued that "causality" was a perceptual a priori with which we process noumena (= the world we can never acess directly). Biolinguists (such as Maturana) on the other hand would argue that "causality" resides in acquired language rather than in the "perceptual system" since the latter is itself an object of "perception". (LAST SENTENCE EDITED)

I think that Derrida used the problem of an infinite regress in definitions as an example of deferred meaning (I think the French term he coined was "differance.") A given object ("sign") is linked to other objects ("signifiers"), and the process of going from one to the other proliferates infinitely, so that you never arrive at a concrete "meaning."

Some theorists interpret this as a problem and say that it is never possible to really know anything or express anything meaningful in language. Others have thought that, even if there is no absolute "meaning" bestowed on the sounds that we have chosen for certain words, there still do exist reasonably steady historical contexts, so that, even if you can't learn English just from a dictionary, it is possible to immerse yourself in the language and learn how to use the arbitrary but still historically real system that is in place. This is why you get such divergent views on the subject: Theoretically, it is hard to justify meaning as anything other than merely arbitrary or true by definition. On the other hand, language seems to be extremely useful nonetheless, which suggests that calling it purely "tautological" is a misrepresentation.

so, if the concept "causality" grew out of our system of language rather than direct sense perception, that means that "causality" is a fiction.

Human beings have to co operate more than some would like us to believe, Basically I consent not to question the absurdity of what you say. This helps us in evolutionary terms as co operation is necessary for the species to survive. Basically we take an axiomiatic approach, in that we all learn what "red " is. Whether we all percieve the same redness, is debatable. Fresco is probably more familiar with Wittgenstein than me: I can safely say I read it and didn't understand it. He also helpfully recanted everything he said.
I think we have to take a reasonable "kantian" approach. Basically communication is always going to be a compromise. It involves two paradigms, hopefully they will coincide. Of course we can always prevent this.
My argument is basically come on people we all know what it is. Lets just love it.

Both "causality" and "an external world" in which it operates could be considered "constructions of human observers". But "fiction" is also a concept within such a world, so it might be better to use the phrase "philosophically naive view of reality" rather than "fiction".

Like all relationships, this one is more complicated. Nothing can truly exist until we have shared the knowledge of its existence. This interpretation and sharing is an intimate act of love, Each one of us breathes life into the barren void.
But language also creates illusion and delusion, so we are at that the same time giving life to reality, and creating a fiction. This fiction in turn becomes reality as we are sharing it. An endless cycle of fraud that is the essence of truth..

An endless cycle of fraud at one level that is the essence of truth at another level

I think you need to add the italics or your comment becomes cryptic.

Thanks: for your comment: I find the accusation of ambiguity has made me chuckle. You have added a NEW LEVEL to the concept of irony and satire.