Completed.

quick-reference/chapters/syntax.tex

@@ -1235,6 +1235,34 @@ \subsubsection{Applications}
 	(_ <: harmonizer :> _)
 \end{lstlisting}
 
+\subsubsection{Abstractions as routing blocks}
+
+Usually, lambda abstractions are supposed to be applied on arguments, using beta-reduction in Lambda-Calculus. But unapplied lambda abstractions can possibly be used in languages that treat them as first-class values\footnote{\url{https://en.wikipedia.org/wiki/First-class_function}}, named closures when they are lexically scoped.
+
+
+
+
+In Faust, a proper semantic has been given to unapplied abstractions as results, which are actually routing circuits, internally called \textit{symbolic boxes}, with their arguments named \textit{slots} and corresponding to inputs. This is a convenient way to work with signals by \textit{explicitly naming them}, to be used in the lambda abstraction body \textit{with their parameter name}. 
+
+For instance a stereo crossing block written in the core syntax:
+
+\begin{lstlisting}
+	process = _,_ <: !,_,_,!; 
+\end{lstlisting}
+
+can be simply defined as:  
+
+\begin{lstlisting}
+	process = \(x,y).(y,x); 
+\end{lstlisting}
+
+which is actually equivalent to: 
+
+\begin{lstlisting}
+	process(x,y) = y,x; 
+\end{lstlisting}
+
+
 \subsubsection{Pattern Matching}
 Pattern matching rules provide an effective way to analyze and transform block-diagrams algorithmically.
 \begin{rail}

src/manual/syntax.md

@@ -1198,9 +1198,9 @@ Whenever the Faust compiler find an application of an abstraction it replaces th
 
 Usually, lambda abstractions are supposed to be applied on arguments, using beta-reduction in Lambda-Calculus. But unapplied lambda abstractions can possibly be used in languages that treat them as [first-class values](https://en.wikipedia.org/wiki/First-class_function), named closures when they are lexically scoped.
 
-In Faust, we choose to give a proper semantic to unapplied abstractions as results, to become routing circuits, internally called *symbolic boxes*, with their arguments named *slots* and corresponding to inputs. This is a convenient way to work with signals by *explicitly naming them*, to be used in the lambda abstraction body *with their parameter name*. 
+In Faust, a proper semantic has been given to unapplied abstractions as results, which are actually routing circuits, internally called *symbolic boxes*, with their arguments named *slots* and corresponding to inputs. This is a convenient way to work with signals by *explicitly naming them*, to be used in the lambda abstraction body *with their parameter name*. 
 
-For instance a stereo crossing operation coded in the core syntax:
+For instance a stereo crossing block written in the core syntax:
 
 <!-- faust-run -->
 ```