6

As in differential form, modular form, quadratic form?

I'm sorry if this is a really silly question.

Gerry Myerson
  • 179,216
HmmmBeee
  • 148
  • 4
    "Form" doesn't really mean anything on its own. It's a historical label that got attached to a few things and then got attached to a few other things by analogy. Forms are usually like functions, but not quite, or something. I wouldn't worry too much about it. – Qiaochu Yuan Jul 20 '16 at 07:36
  • It would be interesting to track down specific first occurences (of "differential form", "modular form" and the like). For example, in Classical Invariant Theory, "form" more or less meant "homogeneous polynomial". I'm ready to believe that this meaning was influential in the naming of differential forms, or modular forms, but a serious historical inquiry would be necessary to establish that... – PseudoNeo Jul 20 '16 at 09:36

2 Answers2

7

A main possibly non-intuitive usage of "form" is as a somewhat particular type of map/function.

Traditionally, the word function was used in a more restrained way and it was mainly used for real and complex functions, only.

For example, classically in (real) functional analysis one would have:

  • A function would be a map from $\mathbb{R}$ to $\mathbb{R}$.

  • An operator would be a map from a space of functions to a space of functions. Example: the map "derivative" so $f \mapsto f'$. The term differential operator is still very common.

  • A form would be a map from a space of functions to $\mathbb{R}$, often a linear one. Example: definite integral, $f \mapsto \int_0^1 f(t)dt$.

It is still common in this context and in linear algebra that forms (linear, bilinear, etc) map from a (vector) space to the reals (more generally the scalar field).

Thus, a form is often a map from a 'complicated' domain (often some space) to a simpler co-domain (typically a field).

The terminology differential form also goes under this umbrella. And, while slightly less clear I'd argue quadratic form (and alike), too.

That said, there are altogether different usages of the word "form" too though, as mentioned in the other answer, notably in the compound "normal form." But there the usage seems more in line with a common sense dictionary meaning of the word. When there are several different yet equivalent ways to express something, then you can write it in one way or in another way, in one form or in another form.

But if the "form" is naturally seen as a map, frequently the co-domain will be 'simpler' than the domain.

quid
  • 42,135
3

I'm not a Mathistorian, but... Likely it originally meant its English meaning of "appearance", and it still does in most usages. Quadratic forms have a very specific appearance, namely a homogenous quadratic polynomial. Modular forms are functions satisfying a certain form of equation and some other conditions. Conjunctive/disjunctive/Skolem normal forms are a particular class of first-order formulae with a certain syntax. Chomsky normal forms and Backus-Naur forms are descriptions of context-free grammars with certain syntactical restrictions on the production rules. And so on...

user21820
  • 57,693
  • 9
  • 98
  • 256
  • 1
    "form" comes from the Latin "forma", meaning "shape". – Alex M. Jul 20 '16 at 08:42
  • @AlexM.: What's the difference? Shape is the external appearance, namely how it appears to us. – user21820 Jul 20 '16 at 08:46
  • I only wanted to emphasize the etymology of the word; what I've said complements, not contradicts, what you've said. – Alex M. Jul 20 '16 at 09:16
  • @AlexM.: Oh I see. I thought you interpreted "form" differently, because there are philosophies which use the word "form" in more abstract senses some of which are specifically not about appearance. – user21820 Jul 20 '16 at 09:35
  • @AlexM.: It is perhaps worth noting that it is not uncommon to read phrases like "Consider a sum of the following shape...". I think these occurrences lend anecdotal credence to the idea that names like quadratic form, modular form, differential form, normal form all stem from the fact that such ideas were discussed before a completely rigorous theory was known. After all, there is not much point in defining "modular form" if you haven't come across enough examples to know what the most important properties of such a function should be, so you just say "We consider a sum of the form...". – Will R Sep 19 '16 at 22:10