17
$\begingroup$

It seems like Gödel didn't use the letter $L$ for his model before his book "The Consistency of the Axiom of Choice and of the Generalized Continuum-Hypothesis with the Axioms of Set Theory", which is probably the first place it got used.

Do anyone of you know why he used the letter $L$? It does seem like a bit ad hoc in the book, where he names some functions $J_i$, then some other functions $K_i$ and then ends up defining $L$. But why $J_i$ then?

(I'm sorry if this is not suited for MO)

$\endgroup$
6
  • 10
    $\begingroup$ the sequence $J,K,L,\ldots$ seems quite natural to me, I would not attach any meaning to it $\endgroup$ Jul 22, 2016 at 6:00
  • 3
    $\begingroup$ A lot of things are ad hoc, and they become standard. $\endgroup$
    – Asaf Karagila
    Jul 22, 2016 at 7:54
  • 1
    $\begingroup$ Perhaps this somehow another instance of the phenomenon whereby Grothendieck used G and K (the first and last letters of his name) for the covariant and contravariant versions of K-theory. $\endgroup$ Sep 22, 2021 at 12:11
  • 12
    $\begingroup$ Maybe it is a universe you wouldn't want to live in, a bloody 'ell. $\endgroup$
    – Ben McKay
    Sep 22, 2021 at 12:38
  • 3
    $\begingroup$ @M.Logic Could you provide a reference or citation for Gödel's believing that $V=L$ should be true? I had heard the opposite. $\endgroup$ Sep 22, 2021 at 15:16

2 Answers 2

14
$\begingroup$

I heard from Kai Hauser that the letter $L$ comes from "law", and it is because the model is constructed using some laws.

$\endgroup$
6
  • 14
    $\begingroup$ Only Gödel was Austrian and his work in german. The german word for law is "Gesetz" so that by your logic $L$ should have been named $G$. $\endgroup$ Jul 22, 2016 at 9:58
  • 5
    $\begingroup$ Let me mention at least that Kai Hauser also is German. $\endgroup$ Jul 22, 2016 at 12:15
  • 21
    $\begingroup$ @JohannesHahn If Gödel's paper was in German, then I could accept your remark, but the fact is that his paper is in English. $\endgroup$ Jul 22, 2016 at 13:06
  • 14
    $\begingroup$ @JohannesHahn It might be considered inappropriate for someone whose name begins with G to name his invention G. $\endgroup$ Jul 22, 2016 at 15:47
  • 10
    $\begingroup$ If this were the explanation, one would expect the word “law” to appear somewhere in Godel’s writing on the subject — does it? $\endgroup$
    – user44143
    May 5, 2018 at 1:55
24
$\begingroup$

So far as I know, there are four interpretations why Gödel uses $\mathrm{L}$ to denote the constructible universe.

(1) By Paul Bernays (see page 64 in Halbeisen 2017), Gödel originally used the old German script capital C in old German script to denote the constructible universe, where enter image description here is a capital $\mathrm{C}$. Although the appearance of capital C in old German script is very like the German script $\mathfrak{L}$ for $\mathrm{L}$, in fact it's one kind of the German script $\mathfrak{C}$ for $\mathrm{C}$. Furthermore Bernay's words don't show why Gödel uses $\mathrm{L}$ to denote constructible universe in 1940 (in fact he had introduced the constructible universe in 1938 which is a proceeding paper). So I don't think Bernays' interpretation is suitable (as desired).

(2) By Kai Hauser, the letter $\mathrm{L}$ comes from "law" since the model is constructed using some laws. But in 1940, Gödel didn't mention any law and the word "law" at all. Furthermore, if so, any other model should be named $\mathrm{L}$ since they are also constructed using some other laws. So I don't think Hauser's interpretation is suitable either.

(3) By someone (I don't remember who said this first), since von Neumann's universe is the first model of $\mathsf{ZFC}$ which is denoted by $\mathrm{V}$ and the constructible universe may be the second model of $\mathsf{ZFC}$, to distinguish from and show relations to $\mathrm{V}$, Gödel extends the angle of $\mathrm{V}$ to 90,rotates it down to the horizontal line, and then makes it become $\mathrm{L}$. Of course, it is a joke, and can't be suitable either.

(4) In my opinion, the only suitable interpretation for Gödel's using $\mathrm{L}$ to denote the constructible universe is that $\mathrm{L}$ is after $\mathrm{J,K}$ in order since Gödel defines the constructible universe $\mathrm{L}$ after defining some functions $\mathrm{J}_\mathrm{i}$ and $\mathrm{K}_\mathrm{i}$ for which beginning with $\mathrm{J}$ is possibly because almost all the letters before $\mathrm{J}$ have been used before introducing the constructible universe in 1940 (and in particular he uses $\mathfrak{Cls}(\mathrm{A})$ related to $\mathfrak{C}$ for $\mathrm{C}$ to denote that $\mathrm{A}$ is a class and $\mathrm{Y}\,\mathfrak{Con}\,\mathrm{X}$ to denote that $\mathrm{Y}$ is connex in $\mathrm{X}$). In fact in 1940, Gödel's usages of many symbols are very random, for example he uses $\mathfrak{A}$ to denote some operation and $\mathfrak{B}$ to denote some notion without any sufficient considerations.


By the way, there are also four interpretations why people use $\mathrm{V}$ to denote the cumulative universe of all sets, or the class of all sets.

(1) Peano used uses $\mathrm{V}$ for "verum", i.e., the propositional truth constant for true, and its upside down form $\Lambda$ for "falsum", i.e., the dual propositional false constant, on page VIII in 1889. He furthermore uses $\mathrm{V}$ to denote the class of all individuals, and $\Lambda$ to denote the class containing no individuals on page XI in 1889 in Latin. The following words improved from comments of @EmilJeřábek may show why Peano made such uses.

On page VIII, Peano uses $\mathrm{V}$ for "verum", i.e., the propositional truth constant for true (that would nowadays be denoted $\top$ or 1), and its upside down form $\Lambda$ for "falsum", i.e., the dual propositional false constant (that would nowadays be denoted $\bot$ or 0), and also, $-$, $\cap$, $\cup$ (that would nowadays be denoted $\neg$, $\wedge$, $\vee$) for the propositional connectives. Then, on pages X–XI, he uses the same symbols $-$, $\cap$, $\cup$ to denote the corresponding operations on classes, and thus $\mathrm{V}$, $\Lambda$ becomes respectively the class of all individuals and the class containing no individuals.

(2) Whitehead and Rusell used $\mathrm{V}$ to denote the class of all objects, which is defined as $\mathrm{V}=\{x\mid x=x\}$ in modern notations, on page 229 in 1910. Whitehead and Rusell also used $\Lambda$ to denote the null class on page 229 in 1910. The following words improved from comments of @Goldstern may show why they made such uses.

$\mathrm{V}=\{x\mid x=x\}$, i.e., $\mathrm{V}=\{x\mid$ $x=x$ is true (verum)$\}$.

After them, people began to use $\mathrm{V}$ to denote the class of all individuals (objects). And nowadays there may be two more interpretations. While I don't know who gave those first.

(3) $\mathrm{V}$ is the v in von Neumann. This is because the cumulative universe is also called von Neumann universe which was first claimed by Gregory H. Moore in 1982 because existence and uniqueness of the general transfinite recursive definition of sets was demonstrated by von Neumann for both Zermelo–Fraenkel set theory in 1928 and his own set theory on pages 745–752 in 1928a. While in fact the first publication of the von Neumann universe was by Ernst Zermelo in 1930. Furthermore, the $\mathrm{V}$ notation was not used by von Neumann in his 1920s papers.

(4) $\mathrm{V}$ is the shape of the picture of the cumulative universe.

$\endgroup$
18
  • 7
    $\begingroup$ Peano’s notation at least makes perfect sense: on p. viii, he uses V for “verum”, i.e., the propositional truth constant for true (that would nowadays be denoted $\top$ or $1$), and its upside down form Λ for the dual propositional constant “falsum” (i.e., $\bot$ or $0$). Also, $-$, $\cap$, $\cup$ denote the propositional connectives we write as $\neg$, $\land$, $\lor$. Then, on pp. x–xi, he uses the same symbols ($-$, $\cap$, $\cup$, V, Λ) to denote the corresponding operations on classes: thus, V becomes the class of all objects, $\cap$ is intersection, etc. $\endgroup$ Sep 22, 2021 at 15:18
  • 7
    $\begingroup$ How does it not explain that? V is used to denote the class of all objects because in general, class operations are denoted by the same symbols as the corresponding connectives. This is just a special case with a nullary connective. $\endgroup$ Sep 22, 2021 at 15:26
  • 6
    $\begingroup$ I fail to see how your opinion on what is good or bad use is relevant to the question how the terminology and notation emerged. But anyway, this does not matter the slightest. If you do not want to call the elements of the domain of discourse objects or individuals, that’s fine: call them by whatever word you prefer instead, say crocodiles. Then Peano’s usage explains why V denotes the class of all crocodiles, and in axiomatic set theory, crocodiles = sets. $\endgroup$ Sep 22, 2021 at 15:49
  • 5
    $\begingroup$ @M.Logic Verum is the predicate true on all objects. So if you are considering classes singled out by predicates, the predicate Verum singles out the class of all objects. $\endgroup$ Sep 22, 2021 at 16:08
  • 6
    $\begingroup$ Rephrasing EmilJeřábek's and მამუკაჯიბლაძე's comments: $V=\{x: v\}$, where $v$ abbreviates "verum". That looks right. $\endgroup$
    – Goldstern
    Sep 23, 2021 at 7:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.