diff --git a/source/basic.tex b/source/basic.tex
index a78628ca87..4e1fdf64e9 100644
--- a/source/basic.tex
+++ b/source/basic.tex
@@ -5559,6 +5559,24 @@
 the object and value representations and accuracy of operations
 of floating-point types are \impldef{representation of floating-point types}.
 
+\pnum
+A floating-point type shall at least represent a subset of rational numbers.
+Depending on the implementation-defined value representation for the type,
+it may additionally represent the non-finite values
+\begin{itemize}
+\item infinity,
+\item a set of quiet ``Not a Number'' values, and
+\item a set of signaling ``Not a Number'' values.
+\end{itemize}
+For any of the above (including zero),
+a floating-point type may either represent a single value or
+two distinct values with negative and positive sign.
+\begin{note}
+A floating-point type which adheres to \IsoFloatUndated{}
+is capable of representing a negative and positive variant
+of all the above\iref{numeric.limits.members}.
+\end{note}
+
 \pnum
 The minimum range of representable values for a floating-point type is
 the most negative finite floating-point number representable