Problem with falsifiability, why is psychoanalysis scientific?

I’ve heard now too many times how psychoanalysis is unscientific and I wanted to express my thoughts on the issue. Especially why falsifiability is not a good indicator whether or not a thing is scientific or not.

That fucking word (falsifiability) keeps getting thrown around like it's the ultimate gatekeeper, you know, this bright line that separates the real scientists from the cranks, but it's bullshit in practice and philosophy of science shows it clearly.

Criteria like falsifiability, replicability, objectivity, reliability, validity and internal/external validity do matter, sure. However, Philosophy of science (e.g., Kuhn, 1970; Lakatos, 1970) shows no single bright line like falsifiability, that separates science from non-science. Science is inductive, involves confirmation alongside disconfirmation and so on and so on.

Strict falsifiability fails in practice. It excludes or struggles to accept fields like evolutionary biology, cosmology, geology and parts of psychology, even economics. And yet it lets through false crank claims as “scientific”.

For example, phrenology from the 19th century is falsifiable, measures skulls systematically and correlates independent personality assessments and life outcomes blah blah blah and is still definitely a pseudoscience.

Empirical testing: Dozens of meta-analyses (e.g. Shedler, 2010) show psychodynamic therapy is effective for depression anxiety and so on. It is superior to inactive controls.

Modern studies use standardized manuals, fidelity checks, outcome measures, meta-analyses, neuroscience (neuropsychoanalysis), and process-outcome studies

A study or application can lack methodology, have overinterpretation without invalidating the field, true of early psychology, medicine, etc. Psychoanalysis has self-corrected: Abandoned/refined unfalsifiable or contradicted elements and adopted empirical standards where feasible.

Why over-relying on falsifiability to label something unscientific is problematic: it dismisses fields with heuristic/explanatory value, clinical/practical success, and incremental empirical support.

Many "unfalsifiable" elements in historical form are now testable via proxies (implicit measures, therapy outcomes). Dismissing it ignores its profound influence on psychology, psychiatry, literature, and culture, plus ongoing empirical validation as an evidence-based treatment per updated criteria.

In summary I would say that psychoanalysis (at least parts of it) count as a scientific field and that it is scientific, because it does the main things good science does: It creates ideas and predictions that can be checked against real life (for example “early childhood experiences shape how people relate to others when they are older”). It tests those ideas with actual evidence and when evidence shows something doesn't hold up it updates and changes its idea rather than ignoring stuff. It also connects and works together with today's proven findings from psychology and neuroscience.

And that's why all this "unscientific" bullshit isn't correct. Psychoanalysis evolves, it gets tested, it delivers results, while people are throwing around falsifiability like a magic wand. Fuck that.

References:
Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the growth of knowledge (pp. 91–196). Cambridge University Press.Kuhn, T. S. (1970). The structure of scientific revolutions (2nd ed.). University of Chicago Press.