“The singer sounds exhausted.”
Maybe.
What did you actually hear?
Shortened phrases? Audible breaths? A narrower dynamic range? Unstable pitch? Or simply a dark timbre you associate with fatigue?
And even if those features are present, can they prove the singer was physically exhausted?
Evidence becomes powerful when the claim does not outrun what the evidence can support.
This is the second pillar leg beneath How Music Works | Music Criticism. The first leg, Music Criticism and Description, builds the listening receipt. This article asks what kinds of evidence different critical claims actually require.
Quick Read
Music criticism uses several evidence types: audible evidence from the sound, measurable evidence such as tempo or frequency, representational evidence from scores and notation, comparative evidence from alternative performances or works, contextual evidence from genre and history, and documentary evidence for claims about intention or circumstance. Strong criticism matches claim type to evidence type, distinguishes observation from inference and signals uncertainty when the receipt is incomplete.
claim → required evidence → source → comparison → counter-evidence → calibrated judgment
Not Every Musical Claim Is the Same Kind of Claim
Consider these four statements:
- The recording lasts 4 minutes 12 seconds.
- The chorus feels more urgent than the verse.
- The producer intended the chorus to feel urgent.
- The chorus is more effective than the verse.
They need different receipts.
The first is measurable.
The second is perceptual and interpretive.
The third requires evidence about intention.
The fourth is evaluative and depends on criteria.
A common critical failure is to use evidence appropriate for one claim to certify another.
Evidence Type 1: The Audible Event
The most immediate evidence is the sound itself.
A critic can point to:
- a tempo change;
- a vocal crack;
- a harmonic arrival;
- a drum entrance;
- a change in texture;
- a clipped transient;
- a pause;
- a repeated phrase.
This evidence is strongest when the critic gives a location and uses language precise enough that another listener can find the event.
Evidence Type 2: Measurement
Some musical claims can be measured.
- duration;
- approximate tempo;
- pitch frequency;
- dynamic level in a digital file;
- spectral distribution;
- number of repeated sections;
- release date;
- chart position.
Measurement constrains debate.
But numbers do not automatically produce aesthetic conclusions.
“72 BPM” is evidence. “Too slow” is a judgment requiring a criterion and comparison.
Evidence Type 3: The Score or Notation
A score can show pitch, rhythm, instrumentation, dynamics, articulation and structural relationships.
It can support claims such as:
- the composer marks a sudden piano;
- the exposition repeat exists in the notation;
- the melody is transferred from violin to flute;
- the harmony delays its resolution.
But notation is a compressed representation of musical possibility, not the whole sounded work.
See How Music Works | Notation.
Evidence Type 4: Recording
A recording allows replay.
That changes criticism dramatically because memory can be checked against the stored sound.
Christopher Dingle’s work on recording and criticism in The Cambridge History of Music Criticism traces how recordings became objects of critical comparison in their own right.
Replay makes possible:
- close comparison of phrasing;
- repeated listening to a transition;
- checking whether a remembered imbalance persists;
- comparing masters or mixes;
- measuring features more carefully.
A Recording Is Evidence of a Recording
This sounds obvious.
It is often forgotten.
A studio recording may contain edits from multiple takes, overdubs, pitch correction, dynamic processing and spatial construction.
It can be strong evidence of the released musical object while being weak evidence of what one continuous live performance sounded like.
See How Music Works | Recording & Production.
Evidence Type 5: Live Witness
A live critic witnessed a particular event from a particular position.
The receipt is valuable because it captures what no later recording may preserve:
- hall acoustics;
- audience response;
- stage presence;
- unplanned accidents;
- physical scale;
- the temporal tension of not knowing what comes next.
It is also bounded by seat, memory and one-time exposure.
A critic should not manufacture replay-level precision from a single live hearing.
Evidence Type 6: Comparison
“Fast.”
Compared with what?
“Innovative.”
Relative to what field?
“Unusually dry vocal.”
Unusual inside which production tradition?
Comparison converts vague difference into calibrated difference.
The Reference Set Is Part of the Argument
A critic can make almost anything look extreme by choosing a convenient comparison.
A fair reference set might include:
- other performances of the same work;
- the artist’s previous releases;
- peer works in the same scene;
- genre exemplars;
- historically relevant predecessors;
- works solving a similar musical problem.
Good criticism explains why the comparison belongs.
Evidence Type 7: Genre Practice
Genre supplies expectations.
Repetition may be central in one tradition and a sign of weak development in another.
A rough vocal may index authenticity in one scene and inadequate control in another.
How Music Works | Genre shows that genre is built from sound, community, history, technology and industry, not simply a playlist label.
Evidence Type 8: Historical Sources
A claim such as “audiences originally found this style shocking” cannot be proven by how the music sounds to us now.
It needs historical evidence:
- reviews;
- letters;
- diaries;
- programmes;
- sales records;
- contemporary commentary;
- institutional archives.
The past does not speak through the recording alone.
Evidence Type 9: Artist Intention
Critics often write, “the artist wants us to feel…”
How do they know?
Possible evidence includes interviews, liner notes, manuscripts, production notes or direct statements.
Without such evidence, safer language is:
the arrangement creates… / the effect suggests… / the passage can be heard as…
This keeps effect separate from intention.
Intention Is Evidence, Not a Veto
Suppose the artist says the track was designed to feel chaotic.
That helps establish intention.
It does not automatically prove that the resulting chaos is effective.
The critic can still ask whether the musical means achieve the stated aim and whether the aim itself produces value.
Evidence Type 10: Reception
Charts, streams, ticket sales, reviews and audience testimony can support claims about reception.
They tell us that people paid attention, purchased, attended, replayed or discussed.
They cannot alone establish aesthetic greatness.
Popularity is evidence about reception. It is not a substitute for an argument about quality.
Evidence Type 11: Expert Testimony
A master percussionist may notice timing relationships most listeners miss.
A producer may identify an artefact accurately.
An ethnomusicologist may explain a local performance convention.
Expertise increases the weight of testimony when the expertise matches the question.
It should not replace reasons where reasons can be given.
Authority Must Match Domain
A famous conductor may be authoritative about orchestral rehearsal.
That does not make the conductor automatically authoritative about hip-hop production history.
Prestige is not universal jurisdiction.
Evidence Type 12: The Critic’s Own Response
Your response is data.
You became tense.
You lost attention.
You replayed the bridge five times.
Those experiences matter.
But one listener’s response does not automatically generalise to all listeners.
The next pillar, Music Criticism and Taste, goes deeper into that boundary.
Evidence Has Provenance
Where did the claim come from?
Official liner notes?
An artist interview?
A fan transcription?
A memory from the concert?
An AI-generated summary?
Criticism becomes safer when source and transformation chain remain visible.
Primary Source Does Not Mean Perfect Source
An artist’s interview is primary evidence of what the artist said.
It may not be perfect evidence of what actually happened in the studio.
Memory can fail. Public relations can simplify. Artists can reinterpret their own past.
Source closeness and source reliability are related but not identical.
Counter-Evidence Is Part of the Receipt
You think the performance is too slow.
Listen for what the slow tempo gains.
Inner lines may become audible.
Text may become clearer.
Harmonic tension may become more spacious.
A judgment is stronger when it survives the best evidence for the alternative interpretation.
Absence of Evidence Is Not Always Evidence of Absence
You cannot hear a bassoon in the mix.
Was it not played?
Was it masked?
Was the microphone placement poor?
Was it edited out?
The recording can establish that the bassoon is not perceptually salient. It may not establish why.
Negative Claims Need Care
“There is no development.”
Maybe development occurs in timbre rather than melody.
“Nothing changes.”
Maybe microtiming, orchestration or density is changing below the critic’s current resolution.
Before claiming absence, widen the search.
Uncertainty Is Not Weakness
Useful phrases include:
- on the recording available to me;
- from my seat in the hall;
- within the comparison set I checked;
- the effect may be deliberate;
- I could not verify the production credit;
- on first hearing;
- after repeated listening.
These phrases bound the claim.
False precision does the opposite.
The Evidence Ladder
For many critical claims, use a ladder:
- Hear: locate the relevant event.
- Describe: state what is observable.
- Measure: quantify if useful.
- Compare: establish whether the feature is ordinary or exceptional.
- Contextualise: identify genre, historical or production norms.
- Interpret: explain possible function.
- Test: seek counter-evidence.
- Judge: state the conclusion at a confidence level the receipt supports.
AI Makes Provenance More Important
A language model can produce convincing sentences about a song it has not heard.
That makes fluent prose a poor proxy for evidence.
If a critic—human or machine—did not inspect the audio, score or source required for a claim, the claim should not be presented as observed fact.
Critical authority should come from a visible evidence path, not from confidence of tone.
Historical Criticism Shows Why Receipts Matter
The history of criticism contains famous misjudgments, shifts of canon and changing evaluative frameworks.
A review is a time-stamped judgment written with the evidence and assumptions available to a particular critic.
This is not a defect unique to old criticism.
It is a reminder that criticism is revisable.
The Singapore Case: Evidence Includes Cultural Position
Shzr Ee Tan’s chapter on music criticism in Singapore in The Cambridge History of Music Criticism is useful because it makes critical position visible. Critics operate inside institutions, cosmopolitan aspirations, local scenes and inherited cultural hierarchies.
A critic’s evidence base should therefore include local knowledge when making claims about local value, authenticity or significance.
A 22-Lens Evidence Audit
- Claim: what exactly am I asserting?
- Type: factual, perceptual, interpretive, causal, intentional or evaluative?
- Audio: what audible event supports it?
- Location: where can the event be found?
- Measurement: can any relevant part be quantified?
- Score: does notation add evidence?
- Recording: am I judging the recorded object or inferring live reality?
- Witness: what did I actually experience live?
- Comparison: what fair reference set did I use?
- Genre: what norms belong to this practice?
- History: what contemporaneous evidence matters?
- Intent: do I have a source for the artist’s intention?
- Reception: what can audience data establish?
- Expertise: whose specialist knowledge is relevant?
- Provenance: where did each factual claim originate?
- Freshness: is the source contemporary with the claim?
- Counter-evidence: what challenges my thesis?
- Alternative: what is the strongest rival interpretation?
- Scope: am I generalising beyond the available evidence?
- Uncertainty: what remains unknown?
- Language: does my wording match confidence?
- Revision: what new evidence would change my mind?
Listening Laboratory 1: Four Claims, Four Receipts
Take one recording and write:
- one measurable claim;
- one perceptual claim;
- one interpretive claim;
- one evaluative claim.
Under each, write the evidence needed to support it.
Listening Laboratory 2: The Intent Trap
Find three sentences in reviews that say an artist “wanted,” “tried,” “meant” or “intended.”
Ask whether the review supplies documentary evidence or merely infers intention from effect.
Listening Laboratory 3: Counter-Evidence Pass
Write your strongest negative judgment.
Then spend one complete listening pass looking only for evidence that would weaken it.
Revise the judgment afterward.
For Primary Readers
Teach one distinction:
I heard it / I think it means / I like it.
Children can learn very early that these are different kinds of statements.
For Secondary Readers
Require every paragraph of musical evaluation to include one located example and one explanation of why that example supports the judgment.
For Advanced Readers
Build an evidence matrix separating source type, authority, provenance, uncertainty and claim scope.
This turns criticism into a genuinely auditable argument.
Common Misconceptions
- “If I can hear it, it proves why it happened.” Audible effects do not automatically prove causes or intentions.
- “Numbers make criticism objective.” Measurement can establish features, but evaluation still requires criteria.
- “The artist’s intention settles the meaning.” Intention is one source; effects and receiver interpretation remain separate questions.
- “Popularity proves quality.” It proves reception at a particular scale and time.
- “Expert opinion needs no explanation.” Expertise should increase resolution and reasoning, not eliminate the public evidence path.
Research and Further Reading
- The Cambridge History of Music Criticism — criticism as an evolving evidential and cultural practice.
- Christopher Dingle — Comparing Notes: Recording and Criticism — recording as a repeatable object of critical comparison.
- Shzr Ee Tan — Music Criticism in Singapore — critical authority and cultural position in a Singapore context.
- Stanford Encyclopedia of Philosophy — Aesthetic Judgment — the normative problem behind aesthetic claims.
Frequently Asked Questions
What counts as evidence in music criticism?
Audio, measurement, scores, recordings, live observation, comparison, historical sources, genre knowledge, expert testimony and reception data can all count, depending on the claim.
Can a personal reaction be evidence?
Yes, as evidence of that listener’s response. It becomes weaker when used to make universal claims about all listeners.
Does measuring tempo prove a performance is too slow?
No. It proves the tempo. “Too slow” additionally requires criteria and a relevant comparison class.
Why does provenance matter?
Because readers need to know whether a factual claim came from direct listening, a primary source, a reliable secondary source, memory or inference.
Final Thought: A Review Is Strongest Where Its Claims Can Be Reopened
A critic does not need laboratory certainty.
Music is not reducible to measurement.
But neither does criticism need to float free of evidence.
A defensible musical judgment is not one nobody can challenge. It is one whose evidence path is clear enough that disagreement can become better listening instead of louder assertion.
MUSIC CRITICISM · FOUR PILLAR LEGS
Return to Music Criticism, or continue through Description, Taste and Context. The wider estate lives in the How X Works Hub.