2

As I understand, the general theory of limits of real functions is enough to infer all of the important results of theory of sequence convergence. However, theory of sequence convergence is most commonly presented before.

In this question I want to verify that the reason for this is only pedagogical (and maybe historical) but that in the end the general theory of limits does not requiere it and that it is only presented a as an intruduction to limiting processes

Moreover, I would like to verify that the definition of the limit of a function when its independent variable tends to infinity and its corresponding associated theorems, e.g. that it exists when the function is bounded and monotonic, are enough to trivialize sequence convergence results

The only possible exception is the General Principle of Convergence, for the one I have only found a demonstration that uses Cauchy Sequences. The purpose would be to dispense all use of sequences as they would just be a subset of real fuctions that have a restricted domain (that is, the natural numbers).

This purpose would serve to present limits first, and then treat sequences as a particular simpler case and use their results when strictly necesary, e.g. when truly being limited to a natural numbers domain. This would put enfassis on the independence and completeness of (truly) real analysis methods, because I think we can get benefit from having independent and complete representations of these 2 sets of real analysis methods

For example, it has been shown that algebra of continuous function theorems can be derived from both, algebra of "general" limits theorems and from algebra of limits of sequences (using Sequential Continuity Theorem) Additionaly, this leads to a simmiliar sepparation between series and integrals where series are limits of sequence partial sums and integrals are limits of "function partial sums" (a proper correspondence is out of scope) I think the dichotomy may be deeper from what is seen at first glance

  • 1
    While most of the well-known advanced calculus text, say Rudin, Wade, Fitpatrick, etc, introduced sequence and limits of sequence at first, and use them as a basis to prove many things for functions, there exists books of advanced calculus that introduce limits of functions first, and postpone the treatment of sequences until the place where they should occur. See Ethan D. Bloch's Real Numbers and Real Analysis. The author also stressed and discussed his pedagogical consideration about this on the preface (it can be freely download at Springer). This is a wonderful book, btw. – Eric Oct 03 '21 at 18:10
  • 1
    An aspect of the depth of the dichotomy has been covered in detail on this site, in that sequential continuity of $f:\Bbb R\to\Bbb R$ does not always imply $(\epsilon , \delta)$ continuity without the Axiom of Choice. – DanielWainfleet Oct 03 '21 at 23:59
  • 1
    @DanielWainfleet Interesting. I'm not sure I have noticed that. Do you have some links as a reference? – Eric Oct 04 '21 at 13:55
  • @DanielWainfleet Same question here – Matias Haeussler Oct 04 '21 at 14:56
  • @Eric . I don't remember & I'm poor at searching this site. Try something like "sequential/topological compactness without AC". – DanielWainfleet Oct 05 '21 at 00:54
  • 1
    @DanielWainfleet Almost did not believe you. This one does the job https://math.stackexchange.com/questions/126010/continuity-and-the-axiom-of-choice – Matias Haeussler Nov 07 '21 at 11:13

0 Answers0