diff --git a/lib/DBIx/Class/Manual/Cookbook.pod b/lib/DBIx/Class/Manual/Cookbook.pod
index 36b891e11..1c60acf76 100644
--- a/lib/DBIx/Class/Manual/Cookbook.pod
+++ b/lib/DBIx/Class/Manual/Cookbook.pod
@@ -1183,7 +1183,7 @@ of your application to support a change lifecycle (e.g. DEV, TEST, PROD) and
 the DB schemas are named based on the environment (e.g. database1_dev).
 
 However, one can dynamically "map" to the proper DB schema by overriding the
-L<connection|DBIx::Class::Schama/connection> method in your Schema class and
+L<connection|DBIx::Class::Schema/connection> method in your Schema class and
 building a renaming facility, like so:
 
   package MyApp::Schema;
@@ -1223,13 +1223,13 @@ building a renaming facility, like so:
 
   1;
 
-By overriding the L<connection|DBIx::Class::Schama/connection>
+By overriding the L<connection|DBIx::Class::Schema/connection>
 method and extracting a custom option from the provided \%attr hashref one can
 then simply iterate over all the Schema's ResultSources, renaming them as
 needed.
 
 To use this facility, simply add or modify the \%attr hashref that is passed to
-L<connection|DBIx::Class::Schama/connect>, as follows:
+L<connection|DBIx::Class::Schema/connect>, as follows:
 
   my $schema
     = MyApp::Schema->connect(
@@ -1244,6 +1244,39 @@ L<connection|DBIx::Class::Schama/connect>, as follows:
 Obviously, one could accomplish even more advanced mapping via a hash map or a
 callback routine.
 
+=head2 Relationships with optional foreign key values
+
+Sometimes you want a relationship where the column holding the foreign key
+value is nullable. In the following example, when C<$human> is deleted the
+related C<$cat> continues to exists and its foreign key value is set to NULL.
+See L<belongs_to|DBIx::Class::Relationship/belongs_to> for details on the
+C<< join_type => 'left' >> attribute.
+
+    package My::Schema::Result::Human;
+    # ...
+    __PACKAGE__->has_many(
+        'cats',
+        'My::Schema::Result::Cat',
+        'human_id',
+        { cascade_delete => 0 },
+    );
+
+    package My::Schema::Result::Cat;
+    # ...
+        human_id => {
+        data_type      => 'int',
+        is_numeric     => 1,
+        is_foreign_key => 1,
+        is_nullable    => 1,
+    },
+    # ...
+    __PACKAGE__->belongs_to(
+        'human',
+        'My::Schema::Result::Human',
+        'human_id',
+        { join_type => 'left', on_delete => 'SET NULL' },
+    );
+
 =head1 TRANSACTIONS
 
 =head2 Transactions with txn_do
diff --git a/lib/DBIx/Class/Relationship.pm b/lib/DBIx/Class/Relationship.pm
index 26a07ef0e..634b96611 100644
--- a/lib/DBIx/Class/Relationship.pm
+++ b/lib/DBIx/Class/Relationship.pm
@@ -191,15 +191,11 @@ more info see L<DBIx::Class::Relationship::Base/condition>.
   # To retrieve the plain id if you used the ugly version:
   $book->get_column('author_id');
 
-
-If the relationship is optional -- i.e. the column containing the
-foreign key can be NULL -- then the belongs_to relationship does the
-right thing. Thus, in the example above C<< $obj->author >> would
-return C<undef>.  However in this case you would probably want to set
+If the foreign key column is nullable you probably want to set
 the L<join_type|DBIx::Class::Relationship::Base/join_type> attribute so that
-a C<LEFT JOIN> is done, which makes complex resultsets involving
-C<join> or C<prefetch> operations work correctly.  The modified
-declaration is shown below:
+a C<LEFT JOIN> is done. This ensures that relationship traversal works
+consistently in all situations. (i.e. resultsets involving C<join> or
+C<prefetch> operations). The modified declaration is shown below:
 
   # in a Book class (where Author has_many Books)
   __PACKAGE__->belongs_to(
@@ -209,7 +205,6 @@ declaration is shown below:
     { join_type => 'left' }
   );
 
-
 Cascading deletes are off by default on a C<belongs_to>
 relationship. To turn them on, pass C<< cascade_delete => 1 >>
 in the $attr hashref.