```Discussion of Homotopy Type Theory and Univalent Foundations
help / color / mirror / Atom feed```
```From: Nathan Carter <nathancarter5@gmail.com>
Subject: [HoTT] my first 3 questions about HoTT
Date: Thu, 20 Jun 2019 09:16:35 -0700 (PDT)	[thread overview]

[-- Attachment #1.1: Type: text/plain, Size: 3184 bytes --]

Hello, HoTT community.

I've learned a bit about HoTT in bits of spare time over the past few
years, and have come up with some questions I can't answer on my own.  It
questions, and if anyone in the community here has time to answer them,
then I'll continue with others as needed.  Thank you in advance for any

(Where I'm coming from:  I'm a mathematician; my dissertation was on
intermediate logics, but I haven't focused on logic much for the past 15
years, instead doing mathematical software and some applied mathematics.  I
have a passion for clear exposition, so as I learn about HoTT, I process it
by writing detailed notes to myself, explaining it as clearly as I can.
When I can't explain something clearly, I flag it as a question.  I'm
bringing those questions here.)

Here are three to start.

1. Very early in the HoTT book, it talks about the difference between
types and sets, and says that HoTT encourages us to see sets as spaces.
Yet in a set of lecture videos Robert Harper did that I watched on YouTube
(which also seem to have disappeared, so I cannot link to them here), he
said that Extensional Type Theory takes Intuitionistic Type Theory in a
different direction than HoTT does, formalizing the idea that types are
sets.  Why does the HoTT book not mention this possibility?  Why does ETT
not seem to get as much press as HoTT?
2. When that same text introduces judgmental equality, it claims that it
is a decidable relation.  It does not seem to prove this, and so I
suspected that perhaps the evidence was in Appendix A, where things are
done more formally (twice, even).  The first of these two formalisms places
some restrictions on how one can introduce new judgmental equalities, which
seem sufficient to guarantee its decidability, but at no point is an
algorithm for deciding it given.  Is the algorithm simply to apply the only
applicable rule over and over to reduce each side, and then compare for
exact syntactic equality?
3. One of my favorite things about HoTT as a foundation for mathematics
actually comes just from DTT:  Once you've formalized pi types, you can
define all of logic and (lots of) mathematics.  But then the hierarchy of
type universes seem to require that we understand the natural numbers,
which is way more complicated than just pi types, and thus highly
disappointing to have to bring in at a foundational level.  Am I right to
be disappointed about that or am I missing something?

Nathan Carter

--
You received this message because you are subscribed to the Google Groups "Homotopy Type Theory" group.
To unsubscribe from this group and stop receiving emails from it, send an email to HomotopyTypeTheory+unsubscribe@googlegroups.com.

[-- Attachment #1.2: Type: text/html, Size: 3666 bytes --]
```

```next             reply	other threads:[~2019-06-20 16:16 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-06-20 16:16 Nathan Carter [this message]
2019-06-20 16:37 ` Cory Knapp
2019-06-20 16:39 ` Thorsten Altenkirch
2019-06-20 16:56   ` Michael Shulman
2019-06-20 23:11     ` Nathan Carter
2019-06-21  1:04       ` Michael Shulman
2019-06-20 16:48 ` Ali Caglayan
```

```Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,

Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

switches of git-send-email(1):

git send-email \
```This is a public inbox, see mirroring instructions