I'm a little lost here, but the claim that you can generate a polynomial function from its differential is wrong.
For example, f(x) = 3x + 1 and f(x) = 3x + 2 are different functions, but their derivative is the same: f(x) = 3. You cannot go 'backwards' from 3 to either of these lines, not even around a single point (they're parallel and share no points in common). The lines can be specified without reference to the differential, as I just did.
Also, the claim that the differential is not a number is confusing: if by 'differential' you mean the result of performing the differential operation on some function, then of course it's not a number, it's another function. The result of differentiating is of the same sort as the thing differentiated, it's just of a lower power.
If by differential you mean the infinitesimal, I don't know what people think about it generally, but certainly you don't need to treat it as a number. You seem to be saying you don't want to treat it as an ideal limit, either, but then, I'm not sure what you're proposing instead. — The Great Whatever
Okay, I wanna backtrack here a little because a) we've both misread the passage on generation, because of my out-of-context quote, and b) I wanna deal more precisely with the Deleuzian treatment of the differential, which I wasn't clear enough about. First, the passage on the creation of the polynomial refers not to (re)creating the entire function, but the function
around the point in question: "You are given one point on a function and a sequence of numbers representing the values of the derivatives of the function at that point, and from these numbers, you can reconstruct an approximation of the whole function not just at that one point, but also in an area around that point." So the reconstruction referred to is pretty local rather than global, but the thrust and stake of the passage is that this locality is nonetheless not correlative to a single, particular point: the successive derivatives express the overall behaviour of points
around any one particular point.*
We can bring out the importance of this seemingly trivial point however if we turn again to Deleuze's reading of the calculus. I said originally that "the differential must differ in kind from the numbers that make up the primitive curve" - this was ambiguous and you were right to call me out on this. It's indeed far more precise to say that the derivative of f(x) yields another function f'(x): what I wanted to convey is that on Deleuze's reading, the difference between these two functions is not simply quantitative but rather
qualitative. What does this mean? Negatively, that the differential cannot be a magnitude or a quantity: at the point at which dy/dx = 0/0, the value of the derivative is itself neither zero
nor an infinitesimal. As Sean Bowden puts it, "
dx represents only the cancellation of quantity in general"; instead, Deleuze's argument is that while it cannot be determined in the form of quantity, it can (only) instead be determined in "qualitative form".
And what does it mean that the differential can be determined only in qualitative form? Simply that, as we've said, the derivative is never simply a value that correlates to a single, particular point on a primitive function, but instead defines the
qualitative character of the function around a particular point. In Simon Duffy's words, "the differential relation characterises or qualifies not only the distinctive points which it determines, but also the nature of the regular points in the immediate neighbourhood of these points" (Duffy, "The Mathematics of Deleuze's Differential Logic and Metaphysics"). This is the import of the Aden quote above. Now, the point of this giant mathematical detour is that insofar as the differential is understood as this element of pure quality ('the cancellation of quantity in general'),
it serves as the model for Deleuze's notion of pure relationality. Again in Bowden's words: "even though
dx is totally undetermined with respect to
x, as is
dy to
y [[
dy/dx can only be determined in relation to each other, without each each value is nothing],
since the relation subsists, they are in principle determinable with respect to each other" (my emphasis).
Why is this reciprocal determinability of the differential important to Deleuze? For two reasons: first, not only does it provide a model for pure relationality, but second and even more importantly, this model itself has a distinctive trait that allows Deleuze to set himself against a position that his entire
oeuvre pitches itself against: the idea that what exists prior to individuation is an
indeterminate generality which is then progressively differentiated though limitation or negation (which itself calls for a correlative abandonment of any hylomorphic model of individuation). In other words, Apo's entire metaphysical picture. What's at stake here? It's this: while Deleuze agrees that one must begin any approach to individuation from the perspective of the
undifferentiated (at the point at which there are not yet 'crisply defined individuals', to use Apo's parlance), it is nonetheless a complete mistake to think that this undifferentiated realm is
indeterminate. On the contrary, he will argue that this pre-individual, undifferentiated sphere of being is
entirely determined - and determined precisely in the qualitative form as outlined above: this is it's 'distinctive trait' that I mentioned.
To bring it all together then, the determination of the pre-individual realm means that it is characterized buy the
distribution of singular and ordinary points. And what does this mean? Again, back to the differential: if we accept that the differential characterizes the qualitative behaviour of a primitive function, then one can argue for the 'existence' of two
kinds of behaviours: singular and ordinary. Points with 'singular behaviours' are, as we've said before, things like inflexion points and stationary points (where the value of a gradient changes or equals to zero or infinity); points with 'ordinary behaviours' are those that remain relatively continuous to their neighbouring points. So with the calculus as his model, Deleuze will refer individuation to the manner in which singular and ordinary points are distributed among a series, and from which,
taken together, a primitive function can be generated. Hence Deleuze's affirmation that "the reality of the virtual [the pre-individual] consists of the differential elements and relations along with the singular points which correspond to them […] Far from being undetermined, the virtual is completely determined.”*
--
It's ultimately over the question of the determination of the pre-individual that the debate between me and Apo turns. Apo is unable to recognize - perhaps because he's never encountered it before - the idea of a determinate but undifferentiated realm of the pre-individual. The terminological disputes over the general and the particular, the singular and the universal, and pretty much the rest of it, all turn upon this difference. The idea that the pre-individual is a vague generality is the 'null hypothesis' which Deleuzian metaphysics tests itself against, even as it responds to a similar motivation - which accounts for the closeness and the distance between our respective position. Anyway, sorry for the long reply, but I'm working on two fronts at once - if you haven't noticed already not all of this post is for you TGW - so I don't have to post multiple times.
*See also Gil Morejon's paper, "Differentiation and Distinction: On the Problem of Individuation from Scotus to Deleuze" (on
academia.edu), which helped me clarify alot of these issues to myself (sneak peak: "What we will suggest is that the positive notion of common nature or virtuality, as something
both completely determined and undifferentiated, far exceeds in its explanatory capacities a negative notion of possibility as purely indifferent... This is an important insight because it forces a re-evaluation of the idea of possibility, which was classically understood as the negation of actuality. Hegel’s famous critique, that Schelling’s philosophy of
Indifferenz amounted to ‘the night where all cows are black’, exposes the paucity of such a conception, through which we ultimately are left incapable of accounting for the reality of actual individuals or distinguishing between empirical instances"
(^ i.e. accounting for the singular - this explains why so many of Apo's posts end up being a simplistic 'taxonomy' of being: things and processes only have 'value' in his system to the degree that they correspond to one or another of his pre-established categories - in themselves, they are meaningless, devoid of significance. I wasn't just being snide when I said earlier that the whole edifice is self-referential - it really is).