Libet time and sight-reading

[Sesquitone]

I have mentioned Libet time (the half-second "awareness" delay) before. The Google AI agent makes a good summary of available literature of the affect on sight-reading:

The "Libet time"—the lag between unconscious brain activity and conscious awareness of an intention to act—is highly relevant to the speed and complexity of sight-reading. It suggests that the brain initiates the necessary motor commands unconsciously, allowing experienced musicians to execute complex, coordinated movements faster than conscious thought alone would permit. This makes sight-reading, a rapid visuomotor skill, an excellent real-world example of the principles demonstrated by Libet's experiments. 

Libet's experiment and unconscious action
In his original experiment, neuroscientist Benjamin Libet found that a readiness potential (an electrical buildup in the brain) occurred approximately 550 milliseconds before a person's movement, but the conscious awareness of the "urge to move" came only about 200 milliseconds before the movement. This implies that voluntary actions are initiated by the brain unconsciously, before we are consciously aware of them. 

How the Libet time affects sight-reading
In musical sight-reading, the rapid, unconscious processing described by Libet's work plays a critical role in how musicians translate written notes into physical actions. The brain handles this complex process of visuomotor translation through several mechanisms:
• Larger eye-hand span: Expert musicians read several notes ahead of where they are currently playing, known as the eye-hand span. While the total time available for processing notes is constrained by the tempo, skilled musicians can process more information within the same amount of time. This suggests that their unconscious processing is highly efficient.
• Anticipation and unconscious processing: The brain processes cues from the musical score to predict what comes next. A 2007 Stanford study found that even non-musicians subconsciously form predictions about musical events. In sight-reading, an expert musician's brain likely anticipates upcoming musical events unconsciously, before they are brought to conscious attention.
• Automatic visuomotor translation: Research shows that after training, musicians activate the superior parietal cortex, a region that automatically translates spatial information (the notes on a page) into motor responses (playing a key). This automatic, sensorimotor translation happens without conscious effort, shortening the processing loop and allowing for near-instantaneous execution.
• Unconscious initiation and conscious veto: The Libet findings suggest that the initial stages of movement preparation are unconscious. However, Libet himself proposed that consciousness retains a "veto" power to halt or modify an action before it is completed. In sight-reading, this would allow a musician to unconsciously initiate a motor sequence but consciously correct a mistake or adapt to an unexpected detail in the music in real-time. 

Unconscious processing for experts vs. novices
The unconscious processing observed in Libet's experiment is particularly relevant to the difference between expert and novice musicians. 
• Experts: Have trained their brains to perform the complex visuomotor translations of sight-reading on an unconscious, automatic level. This efficiency explains why they can read complex music at speed.
• Novices: Rely more heavily on conscious, effortful processing to decode musical notation. This limits their speed, as the bottleneck of conscious thought slows the translation from note to action. 

Summary
The Libet time demonstrates that the brain can initiate actions unconsciously before the mind is consciously aware of them. In the context of sight-reading, this suggests that the rapid and fluent performance of expert musicians relies on highly trained, unconscious brain processes that translate visual information into motor commands. The time-delay between brain activity and conscious awareness is not a liability, but rather a necessity that allows musicians to execute complex actions at speeds that conscious, deliberate thought could not achieve. 

[timothy42b]

Good article, even though AI.

I suggest that mostly applies to prima facie sightreading, which is what beginners assume is sightreading, but is only a subset of what skilled readers do.

IMO, most good sightreading is retrieval of memorized patterns rather than prima facie visuomotor. (and the same is probably true for technique)

There are implications for how to improve it.

[Sesquitone]

Good article, even though AI.

I suggest that mostly applies to prima facie sightreading, which is what beginners assume is sightreading, but is only a subset of what skilled readers do.

IMO, most good sightreading is retrieval of memorized patterns rather than prima facie visuomotor. (and the same is probably true for technique)

There are implications for how to improve it.

Yes, the AI robot is getting a lot better, especially on technical subjects where there is not (too much) controversy. And, yes, I am referring to what some people call "sight-playing". And, as opposed to "memorising", the same (Libet time) considerations apply to well-rehearsed playing-from-music. With the trombone, we have the added "decisions" of when (and where) to play "alternate" positions. Similar to alternate fingerings on keyboards (for which there are well-known "rules").

[timothy42b]

And, yes, I am referring to what some people call "sight-playing". And, as opposed to "memorising", the same (Libet time) considerations apply to well-rehearsed playing-from-music.

I didn't explain well. Or, maybe you simply disagree.

You used the piano example so I'll start there; trombone is different but there are similarities.

To play a note on piano, you see the note, calculate a 3 dimensional trajectory to put your finger on it, and play the note. There are a lot of steps - recognizing the note, knowing what key corresponds to it, knowing the coordinates of that key in three dimensional space, knowing what finger needs to hit it, knowing where that finger is at the moment, moving the finger through space accurately.

That is what I call prima facie sightreading, and it is a major hurdle for beginners. Getting a sense of keyboard geometry without looking seems impossible at first.

And that is a tiny subset of what happens with skilled sightreading.

Most sightreading relies on recognizing patterns that have been thoroughly learned and can be instantly retrieved. That is why it tends to be specific to the genre, and why practice sightreading lots of material is insufficient.

IMO of course.

I recall a physiological psychology course back in the 70s where a studies said neurons in the pyramidal motor cortex fire about a second before any motor impulse, which seems similar to the libet idea you posted. And that is why golf and bowling are so different from pickleball.

[Sesquitone]

I didn't explain well. Or, maybe you simply disagree.
I recall a physiological psychology course back in the 70s where a studies said neurons in the pyramidal motor cortex fire about a second before any motor impulse, which seems similar to the Libet idea you posted. And that is why golf and bowling are so different from pickleball.

You explained well, and we are not in disagreement. The point I'm trying to emphasise about Libet time is that, in many activities, there is not enough time to "think about and make conscious decisions" before acting. Success depends on learned ingrained ("automatic") responses. Even then, especially in some sports, as you mentioned, this may not be enough to overcome the Libet time lag—that's why there are so many aces in professional men's tennis. For a successful return, the receiver will (in the "real" world with events happening half-a-second before any humans are aware of them) have hit the ball (in an appropriate way) before he is actually aware of the server completing his serve! The receiver must rely on other cues such as the "body language" of the server. The server, on the other hand, can easily "outwit" the receiver by a late twist of the racquet so as to send the ball on an unexpected trajectory. [Substitute "curve balls" in baseball, "Googlies" in cricket.]

A novice trombone player, on seeing a quarter-note on the fourth line of the bass clef, will have to think "F" (preferably, "F3"); then think "first-position" and (the embouchure equivalent of "third harmonic")—not to mention noted articulation and dynamics—and then decide to play it in time (which happens half-a-second before she is aware of playing it). Additionally, depending on surrounding notes in the phrase, it may be "better" to play this in sixth position. That's a whole lot of "parameters" to think and decide about. Add in several sharps for a remote key and a lot of accidentals, . . . !

For experienced players, all of this is "automatic"—including the selection of "alternate" positions, especially in the upper register where there are many options. [Or in the otherwise "awkward" lower register if you are using a minor-third attachment.]

Then there is the question of "overcoming" Libet time when improvising (e.g. over chord changes in an original piece at a fast tempo). There's no time to think about appropriate arpeggios, chord-scales, and approach notes, and to then decide what to play—to fit the "arc" of the intended phrasing. Whatever you have in fact decided to play, you will have actually played half-second before you are aware of playing it.

[LeTromboniste]

Most sightreading relies on recognizing patterns that have been thoroughly learned and can be instantly retrieved. That is why it tends to be specific to the genre, and why practice sightreading lots of material is insufficient.

Maybe I misunderstand your point, but to me this doesn't disagree with OP, quite the opposite. You're right, most sight reading is not "first sight", because it consists of patterns we do know. But that's precisely the point. The patterns we learn are not only accessed consciously, the brain also accesses them to make the predictions of the signals it will need to send, so that it can get the right movements started before we are even conscious of what we read.

[brassmedic]

A novice trombone player, on seeing a quarter-note on the fourth line of the bass clef, will have to think "F" (preferably, "F3"); then think "first-position" and (the embouchure equivalent of "third harmonic")—not to mention noted articulation and dynamics—and then decide to play it in time (which happens half-a-second before she is aware of playing it). Additionally, depending on surrounding notes in the phrase, it may be "better" to play this in sixth position. That's a whole lot of "parameters" to think and decide about. Add in several sharps for a remote key and a lot of accidentals, . . . !

For experienced players, all of this is "automatic"—including the selection of "alternate" positions, especially in the upper register where there are many options. [Or in the otherwise "awkward" lower register if you are using a minor-third attachment.]

Then there is the question of "overcoming" Libet time when improvising (e.g. over chord changes in an original piece at a fast tempo). There's no time to think about appropriate arpeggios, chord-scales, and approach notes, and to then decide what to play—to fit the "arc" of the intended phrasing. Whatever you have in fact decided to play, you will have actually played half-second before you are aware of playing it.

I looked up Libet's experiments, and it seems to me you are very much overgeneralizing what he found. Libet found specifically that there is some brain activity a split second before a person is consciously aware of their intent to act. There seems to be some debate as to whether this brain activity is the cause of the action as opposed to just a general readiness to prepare for the impending decision, and Libet himself believed the conscious mind still had time to disregard this subconscious brain signal. So to say that everything we do is subconscious, and we are not aware of doing anything until afterwards, is a big stretch. I think there are reflex actions that we perform that are subconscious, but that isn't what Libet was studying. In fact, some have theorized that time does not move as a narrative, but that we remember it that way after the fact for our own convenience. That's a very compelling idea, but again, not what Libet was showing.

As for sightreading, I feel that you are ignoring the fact that one can decide what one is going to play far in advance of the event. Good sight readers are reading ahead and already deciding what they are going to do long before getting to that place in the music. So 500 milliseconds isn't even going to factor in to this. Having said that, I nevertheless agree with Timothy that an experienced musician does not decide, note by note, what to play, but retrieves memorized patterns that are already well ingrained.

[timothy42b]

Golf vs pickleball.

In one you are freed from time - hit the shot anytime you want. (subject to criticisms of slow play)

In the other the ball is moving and bouncing, there is one instant of time and it is determined externally.

Playing in an ensemble is possibly closer to the second case. Even if you have ten minutes of rest before your note, it needs to be on time.

[Wilktone]

I find this sort of thing fascinating. There are examples of how the reality we experience through our senses is essentially a construct our brains put together to cope for delays in sensory processing (and other neurological and physiological quirks). For example, we see the world right side up, but our eyes actually send the visual elements to our brain upside down. But from the standpoint of reading music, I don’t think it’s very useful to consider how our brains flip our vision. It’s something we’re not consciously aware of.

I have mentioned Libet time (the half-second "awareness" delay) before. The Google AI agent makes a good summary of available literature of the affect on sight-reading:

I like playing around with AI too, but I’ve found you can get different answers from different systems and can get a differing answer depending on how you ask the question. For what it’s worth, here’s what ChatGPT said in response to the query, “ Does Libet time have any impact on how musicians sight read or perform music?”

This is a thought-provoking question. At first glance, Libet’s experiments—and the broader issue of “readiness potentials” (RPs) and the timing of conscious intention—don’t obviously map onto the very rapid, practiced, often overlearned acts involved in musical performance or sight-reading. But it’s worth teasing apart what “impact” might plausibly mean, and where the parallels (or limits) lie. Here’s a detailed take:

⸻

What “Libet time” means (and its caveats)

To recap, Libet’s classic finding was that a measurable “readiness potential” in EEG (a slow buildup of neural activity) precedes the reported moment when a subject becomes consciously aware of their intention (the “W-time”) by a few hundred milliseconds.  This has prompted philosophical debates about free will, conscious vs. unconscious initiation of action, and the extent to which conscious intention “causes” movement.

However, there are several caveats and challenges to generalizing from Libet’s paradigm:
   •   Libet’s experiments typically involve simple, spontaneous, self-paced movements (e.g. flexing a finger), in a deliberately laboratory-reduced context.
   •   There is ongoing debate about how to interpret readiness potentials (e.g. whether they reflect genuine motor planning, general background neural fluctuation, accumulation of spontaneous evidence, or preparation/anticipation). 
   •   Some newer models (e.g. Schurger et al.) suggest that the buildup of neural activity pre-movement might be an “accumulator” of spontaneous fluctuations, not a deterministic causal march to action. 
   •   Libet himself conceded that while the initiation might be sub-threshold, conscious veto (“I decide not to go ahead”) might still play a role. 

Given all that, one must be cautious in extrapolating Libet’s findings to complex, high-speed, expert performance.

⸻

What is required in sight-reading / musical performance

To see whether Libet‐style effects could matter for music, let’s list a few relevant features of musical performance (especially sight-reading):
1. Fast, fine-grained motor control: executing sequences of finger/hand movements with precision and timing.
2. Feedforward and anticipation: musicians often anticipate upcoming notes and plan ahead, rather than purely reacting.
3. Sensory feedback, error correction, adaptation: real-time monitoring and adjustments (tempo, articulation, dynamics).
4. Cognitive load / reading / pattern recognition: the act of decoding the notation, interpreting, planning fingerings, etc.
5. Automaticity and chunking: over practice, many motor sequences become automatized or “precompiled.”

Because of these attributes, a lot of the “action” in musical performance probably lies in well-practiced, prediction-based systems and fast feedforward loops, not purely conscious moment-to-moment decisions.

⸻

Possible ways Libet-type considerations could (or be argued to) influence musical performance

Given the disjunction between the lab paradigm and musical performance, one can still explore possible connections or (more weakly) constraints. Here are some speculations / frameworks:

Hypothetical link How it might manifest in music Likelihood / caveats
Preconscious preparation of motor sequences Before one “decides” to play a passage, neural circuits may already be primed or biased toward certain fingerings or motor trajectories. This is plausible: many motor plans may be pre-activated, particularly in well-learned passages. But this is not exactly what Libet measured.
Temporal constraints and reaction limits Because neural preparation has latency, extremely rapid reading and execution leave little margin for conscious deliberation; decisions must be fast or precomposed. Very plausible: in fast sight-reading, there isn’t time for “wait for conscious intention” per note; the musician must rely on anticipatory planning and motor readiness.
Veto / inhibition / error correction Even if an initial motor plan is underway unconsciously, the performer may consciously suppress or alter it (a sort of “veto”) if they detect an error or unwanted onset. This is analogous to Libet’s notion of a veto window. In performance, real-time control and inhibition are critical (e.g. stopping a finger from pressing a wrong key).
Cognitive bottleneck & timing overlap The interplay of reading ahead, motor planning, and sensory feedback may crowd out a “conscious initiation moment” for each individual note — the system may operate more as a continuous pipeline than discrete conscious decisions. This seems quite likely. The brain may not engage fresh conscious intention for every micro-action in performance.
Delay in conscious awareness If conscious awareness of an intention lags behind neural preparation, then what we feel as “I decided to play that note” might already be preceded by preparatory neural activity. Over many notes, this lag becomes negligible relative to continuous performance. The delay (hundreds of ms) is large relative to some musical events, but in continuous performance that lag may blur or become functionally negligible.

So, while Libet’s classic paradigm doesn’t directly map, the general idea that motor preparation can precede conscious awareness is not inconsistent with how skilled performance likely works (where much of the fine timing is delegated to sub-conscious systems).

⸻

Counterpoints and reasons the effect might be negligible or irrelevant
   •   Expert automaticity: Many aspects of musical performance, especially in trained musicians, become automatized; the role of conscious intention per micro-movement is reduced.
   •   Parallel, pipelined control: Musical performance is not a simple “decide then move” sequence—we see parallel processing, overlapping planning, error correction, sensory feedback loops, etc.
   •   Speed vs. introspection: The rapidity of performance likely exceeds the timescales at which conscious intention can reliably intervene.
   •   Lack of empirical evidence: To my knowledge, there is little direct empirical work linking readiness potentials (or Libet-style measurements) with sight-reading or musical performance tasks.

⸻

What empirical evidence or experiments could test a connection

If one wanted to probe whether a “Libet effect” influences musical performance, here are some possible designs:
   •   EEG / MEG during sight-reading: try to measure readiness potentials or motor preparation signals aligned to note onsets, and see if there are neural precursors before the conscious decoding or decision.
   •   Manipulate decision time: compare performance in conditions where a musician has more vs. less time to plan (e.g. slower vs. faster tempo, or with a “go cue” vs. free reading) and look for signs of pre-activation.
   •   Interrupt / veto conditions: require the musician occasionally to inhibit planned notes (e.g. a “skip this note” go signal) and see whether the inhibitory signal can override a motor plan once it has begun.
   •   Compare novice vs. expert: see whether more experienced performers show earlier motor preparation (or tighter coupling of prediction) than novices.

⸻

Summary & verdict
   •   In short: Libet’s classic results don’t directly imply anything strong about sight-reading or musical performance, but the general notion that neural preparation can precede conscious awareness is not in conflict with what we suspect about skilled performance.
   •   In highly practiced, fast, anticipatory tasks like sight-reading, the margin for conscious initiation per note is small or perhaps non-existent; many movements are likely primed in advance or handled by lower-level control loops.
   •   Thus, one might more fruitfully ask: How much conscious control does a performer retain over micro-actions once they are underway? In that sense, a “veto,” inhibition, or adjustment is more plausible than a fresh conscious decision each time.

If you like, I can look up whether any neuroscience or music cognition research has already attempted to measure readiness potentials in musicians during sight-reading or performance, to see whether there is direct evidence. Do you want me to try that?