diff --git a/man7/utf-8.7 b/man7/utf-8.7
index bed5cb9ee..e4bb94d2c 100644
--- a/man7/utf-8.7
+++ b/man7/utf-8.7
@@ -62,7 +62,7 @@ and
 does not have these problems and is the common way in which
 .B Unicode
 is used on Unix-style operating systems.
-.SH PROPERTIES
+.SS Properties
 The
 .B UTF-8
 encoding has the following nice properties:
@@ -120,7 +120,7 @@ characters may be up to six bytes long, however the
 standard specifies no characters above 0x10ffff, so Unicode characters
 can only be up to four bytes long in
 .BR UTF-8 .
-.SH ENCODING
+.SS Encoding
 The following byte sequences are used to represent a character.
 The sequence to be used depends on the UCS code number of the character:
 .TP 0.4i
@@ -170,7 +170,7 @@ code values 0xd800\(en0xdfff (UTF-16 surrogates) as well as 0xfffe and
 0xffff (UCS non-characters) should not appear in conforming
 .B UTF-8
 streams.
-.SH EXAMPLE
+.SS Example
 The
 .B Unicode
 character 0xa9 = 1010 1001 (the copyright sign) is encoded
@@ -186,7 +186,7 @@ encoded as:
 .RS
 11100010 10001001 10100000 = 0xe2 0x89 0xa0
 .RE
-.SH "APPLICATION NOTES"
+.SS "Application Notes"
 Users have to select a
 .B UTF-8
 locale, for example with
@@ -266,7 +266,7 @@ and
 .B ISO 8859
 at all levels as the common character encoding on POSIX systems,
 leading to a significantly richer environment for handling plain text.
-.SH SECURITY
+.SS Security
 The
 .BR Unicode " and " UCS
 standards require that producers of
@@ -285,7 +285,7 @@ version of "/../" or ";" or NUL and overlook that there are many
 ways to represent these things in a non-shortest
 .B UTF-8
 encoding.
-.SH STANDARDS
+.SS Standards
 ISO/IEC 10646-1:2000, Unicode 3.1, RFC\ 2279, Plan 9.
 .\" .SH AUTHOR
 .\" Markus Kuhn <mgk25@cl.cam.ac.uk>