Font ComparisonMason's Way · four candidate typefaces
For review with Micah
Which typeface carries the code marks, and which carries everything else
Four candidates, all already in the project folder. The page is in three parts: what we are suggesting and why, a full specimen of each face, and a side-by-side of the letter pairs that are easiest to confuse. Nothing here is decided. The recommendations are ours to argue for and yours to overrule.
Part one
Suggested split, screen and print proposed
Two typefaces do most of the work. One has to be able to carry the code marks, and only one candidate really can. The other just has to stay out of the way.
Marked reading text, screen and print Andikastrong case
Every word that carries a code mark. Lessons, flashcards, converted worksheets, the mark legend, worked examples.
Why: Andika includes 106 combining diacritics. The next best candidate here has 19. The marks stack above and below letters, so the font has to be able to compose them rather than just display them. More to the point, it is the only one of the four with marks that sit below a letter, and your method needs those. Andika was also drawn specifically for literacy learners, with a single-story a, a single-story g, and a tail on the lowercase l. Part four shows the test.
The cost: Andika is wide. At the same point size it uses noticeably more line length than the others, which matters more in print than on screen. Budget for that in page layouts.
Interface and navigation, screen Inclusive Sansproposed
Buttons, labels, menus, headings, instructions. Anything that is not marked reading text.
Why: it is quieter and narrower than Andika, so the interface takes up less room and asks for less attention. Using a second face for the chrome also does real work: it tells the reader at a glance that the marked text is the content and everything around it is the machinery. If you would rather use one face everywhere, Andika alone is workable, just heavier.
Unmarked body copy in print Atkinson Hyperlegible Nextproposed
Printed parent handouts, resource sheets, conversation guides. Prose with no marks in it.
Why: it is the current Braille Institute release, drawn for legibility rather than for style, and its letterform disambiguation is the most aggressive of the four. That is what you want in a handout somebody reads once, at speed, on paper. It also gives us weights from extra light to extra bold, which is more range than Inclusive Sans offers for print hierarchy.
Hard limit: 15 combining diacritics, none of them below the baseline. It cannot compose the code marks and must never be used for marked text. If a document mixes prose and marked words, set the whole thing in Andika rather than switching mid-page.
Special purpose only Atkinson Hyperlegible Mononiche
Places where a reader has to see one character at a time: a spelling attempt letter by letter, a nonsense word, letter-by-letter feedback.
Why: fixed width forces every character into its own visual slot, which stops letters from grouping into shapes the eye guesses at. That is useful in a spelling exercise and actively unhelpful in a paragraph. Not body copy, ever.
Part one, continued
Body copy settings to test starting points
Typeface is only half of readability. These are the settings we would start from, drawn from common dyslexia reading guidance. They are starting points for testing with real readers, not established fact.
Size18 to 20 px on screen. 12 to 14 pt in print. Never below 16 px on screen.Andika runs large for its point size, so it may need a touch less than Inclusive Sans to look the same.
Line height1.5 to 1.6 for unmarked prose. 1.8 to 2.0 for marked text.Marked text needs the extra room or a stacked mark collides with the line above it. This is a practical constraint, not a style preference.
Line length55 to 70 characters.Andika will reach 55 characters at a narrower column width than the others, so print layouts need wider measures or a smaller size.
AlignmentLeft aligned, ragged right. Never justified.Justification creates uneven word spacing and vertical rivers of white. This is one of the few points nearly every dyslexia reading guideline agrees on.
WeightRegular for body text. Bold for emphasis. Avoid light weights entirely.Use bold rather than italic for emphasis. Italics slant the letterforms and undo some of the disambiguation these faces were drawn for.
CaseSentence case. Avoid all caps for anything longer than a short label.All caps removes the ascender and descender shapes that make word outlines recognizable.
ContrastDark warm grey on a warm off-white, rather than pure black on pure white.The samples on this page use that pairing so you can judge it directly. Some readers prefer a tinted background; worth offering as a setting rather than deciding for everyone.
Letter spacingSlightly open, roughly 0.01 to 0.02em.Test before applying. Andika is already generously spaced and can look loose if we add more.
Word spacingSlightly wider than default for young or beginning readers.Helps word boundaries stay obvious, which matters in the sentence-segmenting exercises.
Reader controlLet the reader change size, spacing, and typeface, and remember the choice.The strongest thing we can do is stop guessing on their behalf. It also settles arguments about defaults.
Part two
Full specimens
Each face at reading size. The counts come from measuring the actual font files in the project folder. Combining diacritics is the number that decides whether a face can carry the code marks.
Andika
proposed for marked textSIL International designed for literacy learners
glyphs 4,234characters 2,660combining diacritics 106weights available 4 plus italics
Reading is not one skill but several working at once. A reader has to hold the sounds of a word in mind, match them to the letters on the page, and keep the meaning of the sentence alive while doing it. When one of those parts is effortful, the others get less attention. Marking the page is a way of handing some of that work back to the eye.
Confusable sets
I l i 1O 0 oB 8S 5Z 2C Gc ea orn mcl dvv wb dp qu vn hi j
Reading is not one skill but several working at once. A reader has to hold the sounds of a word in mind, match them to the letters on the page, and keep the meaning of the sentence alive while doing it. When one of those parts is effortful, the others get less attention. Marking the page is a way of handing some of that work back to the eye.
Confusable sets
I l i 1O 0 oB 8S 5Z 2C Gc ea orn mcl dvv wb dp qu vn hi j
Reading is not one skill but several working at once. A reader has to hold the sounds of a word in mind, match them to the letters on the page, and keep the meaning of the sentence alive while doing it. When one of those parts is effortful, the others get less attention. Marking the page is a way of handing some of that work back to the eye.
Confusable sets
I l i 1O 0 oB 8S 5Z 2C Gc ea orn mcl dvv wb dp qu vn hi j
Reading is not one skill but several working at once. A reader has to hold the sounds of a word in mind, match them to the letters on the page, and keep the meaning of the sentence alive while doing it. When one of those parts is effortful, the others get less attention. Marking the page is a way of handing some of that work back to the eye.
Read that paragraph and then read the same one in Andika. The fixed width that makes a single letter unmistakable is what makes a paragraph tiring. That is the whole argument for keeping this face to spelling exercises and letter-by-letter feedback.
Confusable sets
I l i 1O 0 oB 8S 5Z 2C Gc ea orn mcl dvv wb dp qu vn hi j
The same set in all four faces, one row at a time. This is the view where the decision actually gets made, because a face can look fine on its own and lose badly in a direct comparison. Worth going through row by row on the call and marking the ones where a face fails.
Set
Andika
Inclusive Sans
Atkinson Next
Atkinson Mono
Rows are built from the list you gave: the classic letter and numeral collisions, the pairs that merge at small sizes (rn reading as m, cl reading as d, vv reading as w), the reversal pairs (b d, p q, u v, n h), and the whole words where the problem compounds.
What to look for
I l i 1 is the hardest test. A face passes if the capital I has serifs or a distinct shape, the lowercase l has a tail or a curve, and the 1 has a base.
b d p q is the one that matters most for this method, since reversal is the pattern Micah is teaching around. Look at whether the bowl and stem differ in more than mirroring.
minimum is the density test. If the letters merge into a picket fence, the face is too even.
An honest caveat
The evidence that any particular typeface improves reading for dyslexic readers is thinner than the marketing around these fonts suggests. What is better supported is that size, spacing, line length, and left alignment help, and that letting the reader choose helps most of all. So this comparison is worth doing carefully for the marks, where the requirement is technical and real, and worth holding loosely for everything else.
Part four
Carrying the marks
This is the test that separates the candidates, and the result is more specific than we expected. All four faces handle the common accents that sit above a letter. The gap opens on marks that sit below the letter, and on marks that stack, which is exactly what your method asks for.
Read the badges, not only the letters. A browser quietly borrows a system font when the chosen face is missing a character, so a cell can look perfectly fine while the typeface itself cannot actually produce it. The badges come from reading the font files directly, so they tell the truth where the rendering does not.
Test
Andika
Inclusive Sans
Atkinson Next
Atkinson Mono
Andika is the only face that carries all of these. Inclusive Sans holds up until the marks get unusual. Both Atkinson cuts stop at the baseline: they have nothing to place below a letter, which rules them out for marked text however good they look above it. The totals behind this are 106 combining diacritics in Andika, 19 in Inclusive Sans, 15 in Atkinson Next, and 14 in Atkinson Mono.
What this does not solve
Most of your marks are not Unicode characters at all. The arc, the suffix box, the final stable syllable bracket, the k-back, the q-curve, and the half-long modified macron have no character in any font, so we draw them as overlays regardless of typeface. Choosing Andika does not remove that work. What it removes is the need to hand-draw the marks that are standard, and it gives the drawn overlays a letterform underneath that was designed with room above and below for them.
Part five
What we need from you
Does Andika look right for marked text?This is the one we would like settled, because the marks depend on it and the rest of the build follows from it. If the letterforms feel wrong to you, say so now rather than later. Your eye for this is better than ours.
One typeface or two?Andika everywhere is simpler and more consistent. Andika for marked text plus Inclusive Sans for the interface is calmer and makes the content stand apart. Both are defensible.
Do the students you work with already have a preference?If you have seen children respond better to one of these, that beats every argument on this page.
Do you print in colour or black and white?It changes how we handle the marks on paper, since colour is one of the easiest ways to separate a mark from the letter underneath and it disappears on a school photocopier.
Is there a face you already use in your own materials?If your handouts have a look that families recognize, we should know before we replace it.