Method
How we review records
A score is a summary of an argument, not a measurement. Here is the argument we are trying to make each time.
Music criticism is not measurement and should not pretend to be. What it can do is describe a record precisely enough that you can predict your own reaction to it.
What we try to describe
| Element | Purpose |
|---|---|
| What it is | Genre, lineage, and what the record is in conversation with |
| How it sounds | Production, texture, dynamics, space — described concretely |
| What it is trying to do | The apparent intent, stated as neutrally as we can manage |
| Whether it succeeds | Judged against that intent, not against what we wish it were |
| Who it is for | The most useful sentence in any review |
| Score | A summary of the above, not an independent measurement |
Why "who it is for" matters most
A record that does exactly what it intends, for an audience you are not part of, is not a bad record. Most disagreements about reviews are disagreements about audience rather than about quality, and saying plainly who a record is for defuses most of them.
What we avoid
- Vague sensory adjectives without referents. "Lush" is a reaction. Describe what produces it.
- Scoring against genre expectations we have not stated. If we are marking a record down for not being what we expected, we say so explicitly.
- Novelty as a proxy for quality. New is not good, and familiar is not bad.
- Reviewing the artist rather than the record. Biography is context, not evidence.
The science, and its limits
Perception research helps us describe mechanism — why a texture segregates, why an intro is memorable, why a groove compels. See why music works. It does not tell anyone whether a record is good. Anybody claiming a scientific basis for aesthetic judgement is selling something.
Commercial relationships
We do not take payment for coverage or for a score. Promotional copies are the normal way records reach reviewers and do not create an obligation. Affiliate links are marked. The full policy is in the disclosure.