Just to (maybe, hopefully) clarify: if $columns contains an array ref with your 22 column names and $values the 22 corresponding values you need to call $sth->execute( @{$columns}, @{$values} ) because it expects to get a flat list of parameters (not references to arrays with them). You're giving it the first below; it wants the second:

$sth->execute( [ "col1", "col2", ... ], [ "val1", "val2", ... ] );  ##
+ not working
$sth->execute( "col1", "col2", ..., "col22", "val1", "val2", ... ); ##
+ what it expects
[download]

Oh the cake....yes, I know it's a lie, but it is just so tasty! :`(
Yes, $columns and $values are strings containing comma separated column names and column values I wish to insert into Pg.

I tried these, none of them work:

$sth->execute(@{$columns},@{$values});
$sth->execute(@columns,@values);
$sth->execute(\@columns,\@values);
[download]

I thought $sth->execute($columns,$values); would be closer to passing the flat comma separated string Pg needs.

What is just not clear in the DBD::Pg cpan documentation is how many arguments $sth->execute(); expects or can take, and I can't see a really clear/obvious example of where a variable is used. Most instances just show a hard-coded value...not very helpful.

You are on the right track with Re^2: DBD::Pg - $sth->execute($csv_var) ERROR, but to understand why this doesn't work

I thought $sth->execute($columns,$values); would be closer to passing the flat comma separated string Pg needs.

you need to be aware of the difference between
• a string containing comma separated list of values (in Perl: a scalar)
• a list of values (in Perl: a list or an array)
So, even DBI could use "placeholded" column names, you'd have to execute ( split(/,\s*/,$columns), split(/,\s*/, $values))