Abstract
This essay will examine some rather serious trouble confronting claims that mathematicalia might be social constructs. Because of the clarity with which he makes the case and the philosophical rigor he applies to his analysis, our exemplar of a social constructivist in this sense is Julian Cole, especially the work in his 2009 and 2013 papers on the topic. In a 2010 paper, Jill Dieterle criticized the view in Cole’s 2009 paper for being unable to account for the atemporality of mathematical existents. Cole’s 2013 paper addresses this objection, providing a modification of his 2009 paper allowing for atemporal mathematicalia. An unusual consequence of Cole’s account is that at least some existential claims about mathematicalia used to be false but now have always been true. By examining the semantics of such claims, we demonstrate that social constructivism is in fact, despite Cole’s attempts to rectify matters, incompatible with atemporal mathematicalia. In the course of examining these semantic details, however, an alternative hybrid view of fictionalism and social constructivism emerges. Those tempted by social constructivism, while perhaps disappointed by the negative results of the paper, may be encouraged by how much of their view can be recovered in this alternative account.
This is a preview of subscription content, log in to check access.
Notes
 1.
This is especially true once we adopt the practice of treating mathematicalia like social constructs.
 2.
It has been suggested to me several times that perhaps we could combine methods here: why not interpret (2) as \(S(\phi ,HQ)\)? The problem with this approach is that, when one works out the semantic details, \(\mathcal {M},t\Vdash S(\phi ,HQ)\) if and only if \(\mathcal {M},t\Vdash \lnot H\lnot \phi \wedge HQ\). But then a model of (1) would still require an \(\mathcal {L}^P\)model modelling the sentence \(HQ\wedge \lnot HHQ\), which we’ve already seen is impossible.
 3.
The astute reader will at this point wonder why we should care about what the truth is about the world at time \(v\) from any perspective other than \(v\). What seems relevant is what the natural diagonal \(\mathcal {L}^P\)model hidden inside the realworld \(\mathcal {L}^{2P}\)model gives us. This is an important point which we will return to shortly.
 4.
Thanks to an anonymous referee for helping clarify this passage.
 5.
The reader whose instinctive response to this is to reach for another temporal dimension may wish to read the appendix at this time.
References
Cole, J. (2009). Creativity, freedom, and authority: A new perspective on the metaphysics of mathematics. Australasian Journal of Philosophy, 87, 589–608.
Cole, J. (2013). Towards an institutional account of the objectivity, necessity, and atemporality of mathematics. Philosophia Mathematica, 21(1), 9–36.
Dieterle, J. (2010). Social construction in the philosophy of mathematics: A critical evaluation of Julian Cole’s theory. Philosophia Mathematica, 18(3), 311–328.
Kamp, H. (1968). Tense logic and the theory of linear order. Dissertation, University of California, Los Angeles.
Venema, Y. (2001). Temporal Logic. In L. Globle (Ed.), The Blackwell Guide to Philosophical Logic (pp. 203–223). USA: Blackwell.
Acknowledgments
This work was supported by the National Science Foundation under Grant No. 00006595.
Author information
Appendix: What if we had more time (dimensions)?
Appendix: What if we had more time (dimensions)?
Given Cole was willing to suppose we could collectively declare objects to have whatever temporal profile suited our purposes, I see no reason to suppose he would be unwilling to grant socialconstructivism doublymodal creativity. Thus, Cole could respond to all of the above by claiming that we can “Declare and collectively recognize that institutional facets of reality have whatever [doubly]modal profile best serves our purposes.”
My response to this is roughly as follows: modifying all the above arguments, we could demonstrate that to making of the account that results from allowing sociallyconstructed doublytemporal profiles demands moving to triply temporal logic. Cole could respond to this by proposing we adopt sociallyconstructed triplytemporal profiles, and the back and forth could go on ad infinitum.
Being more explicit, Cole could propose that socialconstructivism’s modal creativity continues to the second temporal variable. Thus, the community responsible for introducing a social construct could declare and collectively recognize that it had whatever doublytemporal profile suited their needs. Nonetheless, we would have to embrace the following:
(3) sounds at least confusing and possibly contradictory, so we need to introduce a third temporal variable to account for the semantics of (3). The arguments from before will carry over, mutatis mutandis, to establish that Cole must choose again between fictionalism and nonatemporal mathematicalia. We could then repeat this process to force a move to quadruplytemporal logic, etc.
Let’s skip to the end of this process and see what the results are. We introduce the language \(\mathcal {L}^{\infty P}\) of infinitely temporal logic for this purpose.
Syntax
Alphabetically, \(\mathcal {L}^{\infty P}\) differs from a base language \(\mathcal {L}\) of propositional logic by the addition of a countable infinity of temporal modal operators \(H_1,\, G_1\), \(H_2\), \(G_2\), etc. Grammatically, \(\mathcal {L}^{\infty P}\) is the closure of \(\mathcal {L}\) after the addition of all these operators. Natural language equivalents of most of these operators are too complex to be worth spelling out.
Semantics
An \(\mathcal {L}^{\infty P}\)model is a triple \(\langle \mathcal {T},<,V\rangle \), where

\(\langle \mathcal {T},<\rangle \) is a poset (partially ordered set—i.e. \(<\) is an irreflexive and transitive relation on \(\mathcal {T}\)) representing the “times” or “moments” together with their ordering and

\(V\) is a function from \(\mathcal {T}^\mathbb {N}=\{f:\mathbb {N}\longrightarrow T\}\) to the set of functions from \({{\mathrm{\mathbf {Prop}}}}\) (the set of propositions) to the twoelement set \(\{T,F\}\).
For \(f\in \mathcal {T}^\mathbb {N},\, V(f)\) is written \(V_f\). In symbols we have
To specify the notions of truth, validity, etc., we use the semantic details given in Fig. 6. Notice the final two clauses in this semantic theory are actually clauseschemes, one instance for each \(n\in \mathbb {N}\).
\(\mathcal {L}^{\infty P}\) as a solution
The observation to make here is exactly the same as the one we made when analyzing \(\mathcal {L}^P\) or \(\mathcal {L}^{2P}\) as solutions. Suppose the modal creativity of social constructivism allows for us to declare and collectively recognize that for absolutely any function \(f:\mathbb {N}\longrightarrow \mathcal {T},\, \mathcal {R},f\Vdash Q\). Then, it seems the sentence
demands we adopt yet another temporal perspective from which to analyze the semantics of Cole’s account. We would thus be forced to adopt the language \(\mathcal {L}^{(\omega +1)P}\), whose semantic theory concerned valuations \(V:\omega +1\longrightarrow \mathcal {T}\). Again, mutatis mutandis, the above arguments can be given, and again Cole will be forced to abandon one of the core elements of his theory or ascend to a yet more complex temporal language in which to state it.
Rights and permissions
About this article
Cite this article
Logan, S.A. The semantics of social constructivism. Synthese 192, 2577–2598 (2015). https://doi.org/10.1007/s1122901506748
Received:
Accepted:
Published:
Issue Date:
Keywords
 Semantics
 Social constructivism
 Philosophy of mathematics
 Temporal logic
 Philosophy of language
 Metaphysics