Aspects of your internal model make a difference to your practice by functioning as precepts. (This is a version of Peirce’s pragmaticism.) This of course includes your practice of inquiry and reasoning, where precepts guide observation or perception. Peirce explains that the subject of a proposition, which acts like an index in directing attention to some singular object, ‘may be a precept by following which a singular could be found’ (EP2:168; follow this link for context).
A scientific model is one from which testable predictions can be deduced in the form of conditional propositions. Strictly speaking, truth belongs to propositions, not to models. We can test our predictions by comparing them with the results of our experiments, but we cannot compare a sign with its object, a word with its meaning, or a message with its source.
We can’t even compare one model with another, unless some ground of comparison exists which amounts to a more comprehensive model.
It is generally admitted that science is fallible, but often the progress of inquiry is expressed in terms of ‘approximation’ to the truth – as if we could step back and measure how close we were to some absolute reality. This in itself is a model of the process of inquiry, incorporating a more or less mathematical diagram: we imagine ourselves (i.e. our consensus) approaching the truth, in the way that geometrical curve approaches an asymptote (i.e. without ever quite arriving at it).
When a theory works better than previous theories, and has been applied successfully in many situations for a long time, we begin to think of it as a ‘law of nature.’ We have no way of knowing whether some other (as yet unimagined) theory would serve equally well, but if the theory in question seems coherent with other established theories, it gradually becomes integrated into our general model of the world.