# How to produce the symbol ȶ in pdfLaTeX?

I coudn't find this symbol in the comprehensive latex symbol list that looks like a t with a curl. Tried to use detexify with no luck either. I don't even know it's name. Can anyone help me? Thanks!

• Table 11 on page 17, \textctt with the tipa package. Jul 20, 2020 at 22:02

See Sergio's answer if you can use a modern Unicode engine like LuaLaTeX (or XeLaTeX).

If you want to access the symbol in pdfLaTeX, it helps to know that this is an obsolete IPA symbol, so you can use the tipa package: It provides the symbol as \textctt (standing for text curly tail t):

\documentclass{article}
\usepackage[safe]{tipa}
\begin{document}
\textctt
\end{document}


Also a general hint if you don't find symbols using detexify: There is a similar site https://shapecatcher.com/ which searches all Unicode symbols.

• I hope shapecatcher is also mining Bitcoin in the background, as detexify used to do. Jul 21, 2020 at 3:02
• None of the two sites finds the symbol, at least the way I draw it. Jul 21, 2020 at 15:08
• @MartinArgerami it often help to try drawing it multiple times. I can assure you that it can be found by shapecatcher (When I tried drawing it, the right symbol appeared at position ~15) Jul 21, 2020 at 15:36

It is the Unicode Character 'LATIN SMALL LETTER T WITH CURL' (U+0236), see https://www.fileformat.info/info/unicode/char/0236/index.htm

If you use XeLaTeX and STIX Two fonts (see https://ctan.mirror.garr.it/mirrors/CTAN/fonts/stix2-otf/StixTwoMath.pdf, page 8), you can get that character by inserting the unicode code/name in the way your editor accepts it.

In Emacs, "Alt+X insert-char [Enter]", then "0236" (without quotes.)

A bit of a fusion of the previous two answers, which declares either Unicode characters for use with PDFLaTeX or tipa-compatible commands for use with LuaLaTeX or XeLaTeX:

\documentclass{article}
\tracinglostchars=2
\usepackage{iftex}

\pagestyle{empty} % format this MWE for TeX.SX
\usepackage[paperwidth=10cm]{geometry}

\iftutex
\usepackage{fontspec}
\setmainfont{Charis SIL}
\DeclareTextSymbol{\textctt}{\UnicodeEncodingName}{"0236}
\DeclareUnicodeAccent{\textsubbar}{\UnicodeEncodingName}{"0331}
\else
\usepackage[T3,T1]{fontenc}
\usepackage[noenc,safe]{tipa}
\usepackage{XCharter}
\usepackage{substitutefont}

\substitutefont{T3}{\rmdefault}{ptm}
\substitutefont{T3}{\sfdefault}{phv}
\substitutefont{T3}{\ttdefault}{cmtt}
\DeclareUnicodeCharacter{0236}{\textctt}
\DeclareUnicodeCharacter{1E6F}{\textsubbar{t}}
\DeclareUnicodeCharacter{02B2}{\textsuperscript{j}}
\fi

\begin{document}

Some linguists use ȶ to represent a sound in Korean whose IPA symbol is ṯʲ.

Some linguists use {\textctt} to represent a sound in Korean whose IPA symbol
is \textsubbar{t}\textsuperscript{j}.

\end{document}


As the only tipa-encoded fonts are the Computer Modern family, Times, and Helvetica, this template selects Times as the IPA serif font, Helvetica as the IPA sans-serif font and Computer Modern Typewriter as the IPA monospace font.

One disadvantage is that combining characters do not work in PDFTeX, due to engine limitations that Frank Mittelbach has said “cannot be realistically overcome.” You must use precomposed characters with PDFTeX.

A more flexible version might check \iffontchar to see whether the current font has a given symbol, and if not, fall back to one that does. Or it might declare them the way the LaTeX kernel integrated textcomp, with \DeclareTextCommand and \DeclareTextAccent for both Unicode and the T3 encoding.