Why 'Every Good Boy Does Fine' Slows Your Sight-Reading

You learned it as a kid. Maybe a teacher wrote it on a whiteboard. Maybe a parent recited it with you until you had it. Every Good Boy Does Fine — E, G, B, D, F, the five lines of the treble clef. And FACE for the spaces: F, A, C, E.

It worked at the time. You could name a note. That was the goal.

The problem is that a decade later, you're still using it — and it's exactly the thing keeping you from reading music in real time.

Disclosure: I built StaffReader, the note-reading app mentioned at the end of this post. I came to this the hard way: after a week of daily mnemonic-flashcard practice, I tested at 6% accuracy on unfamiliar notes at tempo. The method wasn't working. That's what eventually pushed me to research how fluent readers actually navigate the staff — and to build something different.

EGBDF and FACE aren't wrong — they're just a scaffold

To be clear: mnemonics are genuinely useful. For a first-day student who has never seen a staff before, a phrase that maps lines to letters is a real foothold. Written theory tests at slow pace? Mnemonics work fine there too. The critique isn't that they're bad tools — it's that they were never designed to stay up past the first few weeks.

Nobody tells you that part. So the scaffold stays, and the slow retrieval path gets more and more reinforced, and you end up years later still whispering "Every Good Boy..." before naming a B.

What the lookup chain actually costs you

Here's the process you run every time you see a note and reach for the mnemonic:

1. Locate the line or space the note sits on
2. Count up from the bottom (or from the nearest landmark you remember)
3. Recite "Every Good Boy Does Fine" up to that position
4. Map that position to a letter

Four steps. Run consciously. On every note. Every measure.

At slow practice tempo this feels manageable. But music has a beat. A quarter note at 80 BPM gives you three-quarters of a second. At 120 BPM it's half a second. There is no room in half a second to recite a phrase, count to the third word, and confirm a letter. The chain breaks down — not because you're not trying hard enough, but because the chain itself is too slow for real-time use.

LaBerge and Samuels (1974) established the key distinction here: a process that is accurate is not necessarily automatic, and accurate-but-not-automatic processing still consumes conscious attention. When attention is used up on decoding, none is left for phrasing, rhythm, the next measure, or your hands. That's the attention bottleneck the lookup chain creates — and it's structural, not something you drill your way through by using the same method more.

If you want the longer version of this distinction, we wrote a whole piece on the automaticity gap in note reading.

A useful analogy: the Stroop effect

Consider what happens when you see the word "red" printed in blue ink and someone asks you to name the ink color.

You slow down. The automatic, overlearned response — reading the word — competes with the deliberate response you've been asked to produce. That interference is the Stroop effect, first described by J. Ridley Stroop in 1935.

The Stroop result doesn't prove that mnemonics fail — the analogy isn't that clean. But it illustrates the cost of routing through a deliberate verbal step when an automatic response would be faster. The mnemonic is that deliberate verbal step. It adds a layer of processing between seeing the note and knowing it.

Consistent with this, Grégoire, Perruchet and Poulin-Charronnat (2013, Experimental Psychology) built a musical version of the Stroop task: trained musicians had to name a note's printed letter while ignoring where it sat on the staff. They couldn't fully ignore it — an incongruent staff position slowed them down, because reading the position had become involuntary. That's the end state a verbal lookup chain never produces: recognition that doesn't wait for you to recite anything.

Why you're still using EGBDF a decade later

The mnemonic habit persists because it works well enough to feel like progress. You name the note correctly. You don't make errors. Nothing in the experience signals that the method is the ceiling.

What the experience doesn't show you: you could be recognizing that same note in a single step instead of four. The four-step version and the one-step version both produce the correct answer in low-stakes conditions. Only real-time reading at tempo reveals the gap.

This is the same issue the accuracy vs. fluency distinction addresses: accuracy is a lower bar than automaticity, and the two require different practice. Drilling the mnemonic builds accuracy. It does not build automatic recognition. Those are different skills, built differently.

What replaces the scaffold

The mnemonic's job was to give you some path to the note name. What it should give way to is a different kind of navigation entirely — one that doesn't require reciting anything.

Fluent sight-readers don't decode each note from scratch. They navigate from fixed reference points they've made automatic. The landmark method formalizes what experienced readers do intuitively.

The two anchors for treble clef:

• G4 — the second line from the bottom. The treble clef's curl wraps around this line; the symbol was designed to mark it. If you can see the curl, you can find the G line without counting.
• C5 — the third space up. Centrally positioned, easy to orient to once G4 is solid.

Once those two are genuinely automatic — not "I can get there in a second" but they arrive before you consciously reach for them — you read every other treble-clef note as a distance from whichever anchor is nearest. A4 is one step above G4. B4 is one step below C5. D5 is one step above C5. No mnemonic. No counting from the bottom. One comparison.

That's structurally faster, and it scales. Each anchor you make automatic reduces the mental work on all its neighbors. The landmark approach compared against other methods covers this in more detail, and the broader methods comparison puts it alongside every other approach with honest notes on what the evidence actually says.

How to start dismantling the chain

The practical path:

Stop drilling all the notes at random. Any practice session that cycles through the full staff keeps every note in the deliberate-lookup column. Spread thin, notes never move to automatic.

Pick one anchor note and drill it until it's instant. G4 is the right starting point — it's the most structurally marked note on the treble clef, and every other note eventually orients from it. Drill G4 alone until you can name it before you consciously search. That might take a few days. It's worth it.

Add the next note only when the first one is stable. C5 goes next. Then check: does G4 still come instantly with C5 in the mix? If G4 wobbles when C5 appears alongside it, stay there longer. Depth before breadth. That's the method.

This sounds slower than drilling everything at once. In practice it's faster, because you're building recognition that actually transfers to real-time use — rather than reinforcing a lookup chain that works at the flashcard table and breaks down at the instrument.

Use a clock. Gentle speed pressure is what pushes accuracy toward automaticity. Resist the urge to "just look it up" — every time you recite the mnemonic, you reinforce the detour you are trying to retire.

A note on apps

Most note-reading apps work against this by design. They present the full staff from session one, which is exactly the setup that builds and reinforces lookup chains rather than automatic recognition. If you're evaluating apps, the distinguishing question is: does it let you stay on one note until it's solid, or does it force breadth from the start?

StaffReader is built around the landmark sequence. Stage 1 is G4 only. You advance to Stage 2 at 90% accuracy — not after a set number of reps or a timer, but when you've demonstrated the note is reliable. Stages 1–3 are free; the full unlock is a one-time $4.99. StaffReader is launching soon — join the early-access list and I'll send you the link the day it's ready.

EGBDF isn't the enemy. It's a first-day tool that was never supposed to stay up this long. The goal is to take it down, replace it with two anchor notes that respond without thinking, and navigate from there. That's the move from counting to reading.