If you dig into "animal", "dog", and "Spot", they derive from your motivation to identify a linguistic term for something. This is partly for social communication, partly a desire to recognize something unknown as safe.

There are an infinite number of concepts you can create from any set of experiences. So you have to be selective about what is important and what is not. "Selective" and "important" imply an underlying motivation which decides what to select and what is important to you. You can't get away from motivations.

Indeed the very idea of creating a world model implies there is something objective there for you to find and model on. But is there? Are our scientific models objective? Does that mean they are objectively true, and will never change? History shows that isn't the case.

The alternative is that we create scientific models based on what's useful to us (pragmatism). Again, "useful" means "what satisfies a motive", and we're back where we began.