Jackson supports a modest role for conceptual analysis. When one wants to support the modest role of conceptual analysis, one does not seek to make a claim about the way the world is. There is nothing about conceptual analysis that would make us believe that we have special epistemic access to a mind-independent world. Instead, the analysis concerns ordinary concepts, which are upheld by a large number of people. They are 'ordinary' insofar as they are something accessible to non-philosophers, and they are concepts in the way that metaphysicians talk about concepts. How philosophers have access to these concepts is through their own intuitions about the matters being analysed. Typically, the analyst considers all metaphysically or logically possible positions. Once one has done that, they are able to determine what the ordinary concept is.
The modest role of conceptual analysis hedges its bets against folk intuitions. Even if some philosophical theory were true, it would not change the way that folk view the world. Philosophical theories of causation on this view, for example, would amount to better understanding how non-philosophers think about cause and effect or the nature of causation. That some philosophical theory comes along and undermines that folk view would not undermine the folk view. Change, in the folk sense, is something more robust than the modest role can accommodate. The folk view can withstand a withering philosophical attack precisely because it is a widely held view.
If the modest view of conceptual analysis leaves the folk view as it is, then what about the immodest view? Does the immodest view have such a robustness that the folk view is undermined. Jackson writes that the immodest role of conceptual analysis "gives intuitions about possibilities too big a place in determining what the world is like" (p. 43f). Here, we see that the individual philosopher's concepts outweigh that of any general folk view. That folk view is rejected because the philosopher's view is correct. Perhaps a good source for the immodest view is Gary Gutting's What Philosophers Know (2009). In that book, Gutting argues that what philosophers know not only outweighs the average folk view but also that it should outweigh it. The philosopher's view should outweigh the novices because it has special access to the world and that warrants us to believe almost anything that the philosopher says about it.
What is so modest about claiming that change is different for the folk? What makes the immodest role of conceptual analysis a danger is that it stakes a very fundamental claim in an argument for the way the world is. George Bealer, too, has claimed that metaphysicians have special access to a set of possible worlds that makes them less susceptible to any form of cognitive bias.
One might believe that Jackson's modest account should win the day because it keeps hold of the non-philosopher's intuitions in coming up with the best theory of x. Here's where things get tricky for the modest account of conceptual analysis. In the modest account, Jackson claims that the folks' sense of change is more robust than a philosopher's sense of change. We have just seen how Bealer-style or Gutting-style forms of conceptual analysis have failed because they believe that philosophers get at something special, which is over-and-above all other forms of reasoning. That is, the philosopher's sense of change is more robust than that of the folk. So, taking this full circle, to say that some concept is more robust than another is a claim about the way the world is. Since the modest account depends upon the folk view being more robust than the philosopher's, it seems, therefore, that Jackson's own modest role of conceptual analysis accords with the immodest role. Given that the immodest view should be rejected, the modest role should too.
If the modest role accords with the immodest role, then on one interpretation the modest role is identical to the immodest role. The modest account, thus, suffers from the same flaws as the immodest account of conceptual analysis.