I’d argue that Forensic Linguistics is protoscientific instead of pseudoscientific. It lacks the necessary scientific rigour, such as the ability to test its own hypotheses; but the reasoning is usually backed by actual Linguistics, and it’s accurate in a way that protoscience typically isn’t.

The difference is important - a pseudoscience is better thrown away, but a protoscience can be still fixed or at least dismantled for parts.