« previous post | next post »

Three years ago, we looked at the decline in handwriting skills, both in alphabetic languages and with characters:  "Cursive and Characters: Dying Arts".  See also "Japanese survey on forgetting how to write kanji", "The esthetics of East Asian writing", and several posts on "Character amnesia".

Before the advent of computers, cell phones, and other electronic devices which do our writing for us, it wasn't always this way.  Penmanship was a discipline that students practiced assiduously, and calligraphy was an art that vied with painting for compositional excellence and esthetic appreciation.

Typed and printed writing is generally clear for any literate reader, but overt legibility was not always a desideratum for handwriting.

If you want to have some fun, do an image search for '"Thomas Aquinas" handwriting'.  For example, look at this sample on this webpage (the language is Latin).  St. Thomas Aquinas surely must be one of the most egregious offenders in Western history for illegible handwriting; his output seems even more opaque than much of Chinese grass script (草書; M cǎoshū, C cou2syu1, which Wikipedia states should be translated as "cursive script").  (Cf. "semi-cursive script" [also called "running script"], which is less devilishly difficult to decipher than grass script.)

As Stephan Stiller (personal communication) puts it:

One might well think of grass script as the world's first known write-only script.  Note that grass-script is normally for the representation of Literary Sinitic (Classical Chinese), which requires more than a language model for a reader to understand, even in printed form.  A write-only language is a programming language with the reputation that code written in it is difficult to comprehend. The most commonly given examples are APL and Perl; APL is old but Perl is widely used.

I sympathize with those who are frustrated when they try to read the handwriting of St. Thomas Aquinas and Chinese cursive script.  People often complain that my chicken scratches are unreadable, even when I try to make them painfully neat and clear!


  1. Victor Mair said,

    March 30, 2014 @ 3:45 pm

    An old e-mail that Leonardo Boiko wrote to me on October 2, 2013:

    I thought maybe you'd be amused by this: (from )

    This is not true cursive support, of course, just a few shortcuts for common strokes. But I found it interesting how a now-antiquated technique (cursivization) was reused in a modern technological context, for much the same reasons.

    (Of course, as you have often noted, the main input methods in China and Japan are all phonological; drawing characters on your phone is useful mostly in exceptional cases, e.g. when one wants to look up unknown characters.)

  2. Victor Mair said,

    March 30, 2014 @ 3:47 pm

    And this came from John Brewer on May 31, 2013:

    When I saw the subhed “Cursive Writing Makes Kids Smarter” (in the Psychology Today piece linked below) my first reaction was “This seems like the sort of thing myl enjoys debunking as, at a minimum, probably overstating what the underlying published research actually shows.” I’m cc’ing Prof. Mair because I am guessing that some of the arguments made might parallel some of those made in East Asia about the dreadful consequences that would supposedly ensue if students no longer spent a lot of time learning to write hundreds/thousands of different kanji/hanzi by hand.

  3. Y said,

    March 30, 2014 @ 4:10 pm

    I can't imagine how anyone was ever able to read Demotic Egyptian either. Some of it looks to me like an assortment of vertical lines of various lengths.

  4. Stephan Stiller said,

    March 30, 2014 @ 4:25 pm

    Hard to say, because familiarity with the language matters so much. See my comment to a previous LL post (handwriting is most often intelligible only with something like a so-called language model).

  5. Michael Carasik said,

    March 30, 2014 @ 4:52 pm

    And see the following from yesterday's N. Y. Times:

    "In an Era of Squiggles, You Can’t Tell the Players Without a Handwriting Analyst"
    By TYLER KEPNER MARCH 28, 2014

  6. maidhc said,

    March 30, 2014 @ 5:05 pm

    Stephan Stiller was replying to my comment that I can read printed Cyrillic but I have a hard time with cursive. His comment is on target, as I don't speak Russian and I have no idea about the grammar and structure of the language. I can recognize placenames and words that are similar to English. If I knew Russian, presumably I would already have some expectation of what was going to be there. Also, I almost never encounter Cyrillic cursive anyway. It's possible that if I had to read it every day, I'd get better at it.

    My mother learned to read printed Arabic without being able to speak the language, because at one time her job involved ordering books in Arabic. She had to order the books, and when one came, match it up with an outstanding order. Understanding what it meant was not necessary for the most part.

  7. Ellen K. said,

    March 30, 2014 @ 5:42 pm

    I think the baseball autograph article linked by Michael Carasik improperly conflates readability and elegance.

  8. blahedo said,

    March 30, 2014 @ 6:04 pm

    One of the things that makes Cyrillic cursive and italic particularly difficult to read for the Western (i.e. natively Latin-script) dabbler is that there are several letters that look similar or identical to some Latin-script cursive/italic letters but correspond to other Latin-script cursive/italic letters.

    The ones that merely look different aren't too bad, because you see them and immediately know you need to do a mental lookup to remember which letter it is. For instance, the cursive/italic form of г (pronounced /g/) is г, which looks like nothing in the non-italic Cyrillic alphabet nor anything in the Latin alphabet.

    But when a Westerner sees т, they unavoidably process this as an M, but in fact it is the italic form of т (pronounced /t/). The italic form of the Cyrillic letter м (pronounced /m/) is м.

    So it's tricky. Of course, it's nothing that a little practice won't solve.

  9. Avinor said,

    March 30, 2014 @ 7:16 pm

    I always thought I was cheating when I had to employ my own internal language model to be able to make sense of the cursive scribblings which in Sweden is typical of my grandparent generation. Interesting to hear that that is how it actually works.

  10. Matt said,

    March 30, 2014 @ 7:50 pm

    Was Aquinas's handwriting equally illegible to his peers, I wonder? Like, are there contemporary (or near-contemporary) sources saying things like "Aquinas's quinque viae seem profound but I don't understand the part about the purple monkey dishwasher"? Or was it within the bounds of acceptability for that point in space and time and culture?

    (It looks bad from here, no argument, but obviously since we weren't trained in the same milieu that's not really any more meaningful than "these two words in contemporary Language X and Language Y look similar so they must be cognate".)

  11. Harold said,

    March 30, 2014 @ 8:14 pm

    Petrarch complained that the writing of the Scholastics (the followers of Aquinas) was cramped and designed for anything but reading. It was very angular with little or no spacing between the words and used a lot of abbreviations. He and his followers developed italic script, using lower case letters, which was thought to be based on classical models but in fact dated to the time of Charlemagne. This coincided with a large increase in book manuscript production for wealthy collectors that occurred just before the invention of printing. I also think italic script is a nice invention. It is very legible, unlike the loopy copperplate that we were all taught in the USA, and in an ideal world would be adapted to be taught in schools.

  12. Fernando Colina said,

    March 30, 2014 @ 9:36 pm

    Ah APL "A Programming Language." It truly encouraged programmers to be as cryptic as possible. Its special character set made programs written in it look like satanic invocations to the uninitiated. It was also powerful. If you had to use more than one line of APL code to implement even the most complex algorithm you were a newbie or a wimp. Its recursivity was wonderful. You could bury recursivity upon recursivity and it was a thing of beauty.

    Which leads to me to say, it wasn't the cursive, it was the recursive that set APL apart.

    Thank you for reawakening the memories, Victor Mair.

  13. Matthew Stuckwisch said,

    March 30, 2014 @ 10:33 pm

    Isn't (Latin) cursive the recommended script for dyslexics? How is that handled for other writing systems?

    Interestingly, when I worked in Spain, students learned cursive first. Early on, all the printed materials were in cursive, and then bit by bit they went introducing the print styles, though students still had to write in cursive.

  14. Matt said,

    March 30, 2014 @ 10:34 pm

    Adding to Harold's comment above in answer to my own question, Mary J. Carruthers confirms that Aquinas's handwriting was considered so atrocious even at the time that multiple secretaries were assigned just to take dictation from him:

    We know a good deal about the actual procedures that Thomas Aquinas followed in composing his works, thanks in part to the full accounts we have from the hearings held for his canonization, and in part to the remarkable survival of several pages of autograph drafts of certain of his early works. Both sources of material have received a thorough analysis from the paleographic scholar, Antoine Dondaine. Dondaine's work confirmed the existence, alluded to many times in the contemporary accounts, of a group of three or four secretaries who took down Thomas's compositions in a fair hand from his own dictation. The autographs are written in littera inintelligibilis, a kind of shorthand that fully lives up to its name (Dondaine says that the great nineteenth-century editor, Uccelli, lost his eyesight scrutinizing these drafts) for it was not designed to be read by anyone other than the author himself. As Dondaine has reconstructed the process of composing the Summa contra gentiles, an early work for which a number of autograph leaves exist, Thomas wrote first in littera inintelligibilis and then summoned one of his secretaries to take down the text in a legible hand while Thomas read his own autograph aloud. When one scribe tired, another took over.

  15. Mark Mandel said,

    March 31, 2014 @ 12:32 am

    Unfortunately, blahedo's comment depends on the font rendering, and for me the "italics" come out as simply oblique (slanted) versions of the normal letters.

    Fortunately for me, I know the Cyrillic alphabet; and fortunately in general, it doesn't take much to figure out b's point.

  16. rosie said,

    March 31, 2014 @ 2:30 am

    @blahedo Italic Cyrillic is even worse than that. Yes, m is misleading to the Roman-script reader because it looks like an M. But it's also misleading because lowercase Cyrillic T in a roman font is т, so m looks as if it is a different Cyrillic letter.

    To be fair, people not used to the Roman alphabet might be similarly misled by a and a (at least in a true italic font), but Cyrillic has this problem — of letters looking different in roman and italic — with several letters. For example, italic д looks like d, and like g as used in written Serbian.

  17. maidhc said,

    March 31, 2014 @ 3:33 am

    Harold: I learned italic script in school. It's served me well ever since. We had some good course materials that were developed in Britain. We had to do it with a fountain pen too. My school had calligraphy contests and the best examples were put up on the wall.

    I never did learn to write cursive but I've never found any need for it either.

  18. Stephan Stiller said,

    March 31, 2014 @ 3:35 am

    I know that doctors' handwriting has a reputation of being illegible in at least three countries far apart from each other on the world map. This notoriety will come from the abbreviated style used on prescriptions: Doctors and pharmacists have a very particular language model they're applying, and it's not surprising that their writing looks like chickenscratch to the uninitiated. Also they're using plenty of abbreviations (cf next ¶).

    »Harold [= following up to Harold]
    The abbreviations Harold is referring to are called "scribal abbreviations". But it seems like Aquinas' handwriting was independently illegible.

    » John Brewer (via Victor Mair), Ellen K., Matthew Stuckwisch
    It turns out that the benefits of teaching handwriting are not clear. In Germany there is a debate about whether cursive script should be taught at all. According to an article entitled [sic ☺] Schreibschrift, ade? (2010-08-24) on the website of the German newspaper FAZ, there is a lack of empirical research in comparing the three cursives taught in the different Bundesländer (parts of Germany) wrt legibility, speed of writing, and orthography [I assume they mean whether or how much any of them facilitate or inhibit learning orthography; I am not sure why or how a cursive script would factor in]. I think cursive script is not taught at all in the various Chinese locales, but there are books (eg 手写速记教程, 978-7-300-09498-4) for aspiring secretaries teaching this and related skills (stenography etc). The question is what precisely the benefits of a cursive script taught in elementary school (or later, for Chinese) could or should be; in addition to the criteria mentioned above I have encountered at least the following: aesthetics, facilitation of a development of individual styles, development of dexterity (motor skills) [like the Psychology Today article mentioned by John Brewer claims]. Let me also observe how antagonistic to each other some of these goals are: writing legibly, writing with style, and writing fast are surely goals best served by differently designed scripts. I'd say that pedagogues need to first figure out what they want to achieve before proceeding in their argument.

    Similar considerations apply to programming languages: most of them grew historically, and the design goals of ease of writing for the non-expert (learnability), speed of writing for the expert, ease of reading, code safety (by this I mean how hard it is to make difficult-to-detect errors, not security of the compiled program against hacking, which is a related yet different goal) [would that be the analogue to the criterion of "orthography" mentioned in the FAZ article?] go against each other. Note that expressivity is yet another thing, even though it interacts with these goals. In this spectrum, Perl can be described as sort of a stenographic programming language, while Java is verbose (and thus has more built-in safeguards against bugs).

  19. Stephan Stiller said,

    March 31, 2014 @ 4:28 am

    I should clarify that by "teaching handwriting" above I mean "teaching a cursive". Of course Chinese students should be able to handwrite characters. What is (in my understanding) not taught in Chinese schools is any specific cursive way of writing Chinese.

  20. Jan said,

    March 31, 2014 @ 5:09 am

    I remember visiting the US for the very first time in 1980 and being shocked by the childishness of many educated individuals' ( teachers and other professionals) handwriting. An awkward kind of print with uppercase letters used at random.

    So if handwritng has deteriorated I dread to think what it is like now.

  21. D-AW said,

    March 31, 2014 @ 7:55 am

    On "chicken scratches" and "chickenscratch":

    “Chicken scratch”, coined by Shakespeare in 1909?"

  22. Stephen said,

    March 31, 2014 @ 9:40 am

    Jan — "Childish" by what standards? "Awkward" by what standards? If you meant that *you* didn't like the way it looked, or that *you* found it hard to read, it's okay just to say that. I think it's probably more accurate, and probably less inflammatory, to assume that people find a way of writing that works for them. I certainly don't hear people complaining, "Gee, I wish I'd learned to handwrite properly!"

    (I doubt that the upper case letters were used "at random" either–I know some people who invariably upper-case r and n, for instance.)

  23. chris y said,

    March 31, 2014 @ 10:25 am

    (the language is Latin)

    How could they tell, as Mrs Parker famously remarked on another occasion.

  24. Peter said,

    March 31, 2014 @ 6:17 pm

    @Victor Mair
    And I've always thought 草 of 草書 means "careless". Actually, I think it means, or at least implies, "hurried".

    @ Stephan Stiller
    I am too young for it, but I think 草書 was taught in my father's times (he was born in 1933). Like the cursive script for English, it allows for faster writing, but as you said, it also establishes a standard (model) so people will be able to read it. Simplified Chinese is based on many of the standards of 草書.

    草書 is also a recognized style in Chinese calligraphy that is taught to those studying the art form.

  25. Kate Gladstone said,

    March 31, 2014 @ 11:45 pm

    Handwriting matters — but does cursive matter? The research is surprising. For instance, it has been documented that legible cursive writing averages no faster than printed handwriting of equal or greater legibility. (Sources for all research are listed below.)
    Further research demonstrates that the fastest, clearest handwriters are neither the print-writers nor the cursive writers. The highest speed and highest legibility in handwriting are attained by those who join only some letters, not all of them – making only the simplest of joins, omitting the rest, and using print-like shapes for letters whose printed and cursive shapes disagree.

    Reading cursive matters, but even children can be taught to read writing that they are not taught to produce. Reading cursive can be taught in just 30 to 60 minutes — even to five- or six-year-olds, once they read ordinary print. (In fact, now there's even an iPad app to teach how: named "Read Cursive," of course — .) So why not simply teach children to read cursive — along with teaching other vital skills, including some handwriting style that's actually typical of effective handwriters?

    Educated adults increasingly quit cursive. In 2012, handwriting teachers were surveyed at a conference hosted by Zaner-Bloser, a publisher of cursive textbooks. Only 37 percent wrote in cursive; another 8 percent printed. The majority — 55 percent — wrote a hybrid: some elements resembling print-writing, others resembling cursive. When even most handwriting teachers do not themselves use cursive, why mandate it?

    Cursive's cheerleaders sometimes allege that cursive makes you smarter, makes you stunningly graceful, adds brain cells, instills proper etiquette and patriotism, or confers other blessings no more prevalent among cursive users than elsewhere. Some claim research support, citing studies that consistently prove to have been misquoted or otherwise misrepresented by the claimant.

    For instance:

    The much-ballyhooed difference in SAT scores between cursive writers and non-cursive writers is … brace yourself … 1/5 of a point on the essay exam. That's all.

    (Yes, I checked with the College Board — see below for the source info they sent me — because not one of the many, many media that mention the "slightly higher" difference actually states _how_much_"slightly higher" the difference is. The College Board researchers who found the difference note, in their findings that this one isn't statistically significant: in other words, it's so small that it's less than the difference you'd expect if the same person took the same test twice. In fact, it's even smaller than the score differences between males and females taking the SAT.)

    So far, whenever a devotee of cursive claims the support of research (e.g., in the PSYCHOLOGY TODAY blog-piece), one or more of the following things has become evident when others examine the claimed support:

    /1/ either the claim provides no traceable source,


    /2/ if a source is cited, it is misquoted or is incorrectly described (e.g., an Indiana University research study comparing print-writing with keyboarding is perennially misrepresented by cursive's defenders as a study "comparing print-writing with cursive"),


    /3/ the claimant _correctly_ quotes/cites a source which itself indulges in either /1/ or /2/.

    (Phenomena /2/ and /3/ were common throughout the PSYCHOLOGY TODAY blog-piece — as was swiftly noted, once the comment-thread began building, by the commenters there. After several hundred comments had gone back and forth — many from the author of the piece — the later comments by that author showed that had significantly changed his views because of the commenters' queries.)

    What about signatures? In state and federal law, cursive signatures have no special legal validity over any other kind. (Hard to believe? Ask any attorney!)
    Questioned document examiners (these are specialists in the identification of signatures, then verification of documents, etc.) inform me that the least forgeable signatures are the plainest. Most cursive signatures are loose scrawls: the rest, if they follow the rules of cursive all, are fairly complicated: these make a forger's life easy.
    The individuality of print-style (or other non-cursive style) writings is further shown by this: six months into the school year, any first-grade teacher can immediately identify (from the writing on an unsigned assignment) which of her 25 or 30 students wrote it.
    All writing, not just cursive, is individual — just as all writing involves fine motor skills. That is why, six months into the school year, any first-grade teacher can immediately identify (from print-writing on unsigned work) which student produced it.

    Mandating cursive to preserve handwriting resembles mandating stovepipe hats and crinolines to preserve the art of tailoring.


    Handwriting research on speed and legibility:

    /1/ Arthur Dale Jackson. “A Comparison of Speed and Legibility of Manuscript and Cursive Handwriting of Intermediate Grade Pupils.”
    Ed. D. Dissertation, University of Arizona, 1970: on-line at

    /2/ Steve Graham, Virginia Berninger, and Naomi Weintraub. “The Relation between Handwriting Style and Speed and Legibility.” JOURNAL OF EDUCATIONAL RESEARCH, Vol. 91, No. 5 (May – June, 1998), pp. 290-296: on-line at

    /3 Steve Graham, Virginia Berninger, Naomi Weintraub, and William Schafer. “Development of Handwriting Speed and Legibility in Grades 1-9.”
    JOURNAL OF EDUCATIONAL RESEARCH, Vol. 92, No. 1 (September – October, 1998), pp. 42-52: on-line at

    Zaner-Bloser handwriting survey: Results on-line at

    Background on our handwriting, past and present:
    3 videos, by a colleague, show why cursive is NOT a sacrament:



    (shows how to develop fine motor skills WITHOUT cursive) —

    Yours for better letters,

    Kate Gladstone
    Handwriting Repair/Handwriting That Works •

  26. Victor Mair said,

    April 1, 2014 @ 6:58 am


    "And I've always thought 草 of 草書 means 'careless'. Actually, I think it means, or at least implies, 'hurried'."

    The 草 of 草書 does not mean "careless". For some people such as yourself, it may be said to connote "careless" and imply "hurried", but, as zdic correctly informs us, cǎoshū 草书 signifies "grass characters; calligraphy executed with strokes flowing together; Chinese characters written in the cursive hand".

    zdic also tells us that 草 by itself means "grass; straw; thatch; herbs" and that it can enter into some expressions such as cǎoshuài 草率 ("carelessly; perfunctorily"). Perhaps it the function of 草 in this small, specialized group of words that has led you to believe and emphasize that as its main meaning in 草書.

  27. Levantine said,

    April 1, 2014 @ 11:06 am

    Cursive handwriting in English seems a very American concern to me. In the UK, my friends and I were taught what was called "joined-up writing", a rather inelegant simplification of real cursive: This was back in c. 1900.

    Years later, when I discovered that American schools were teaching an altogether more beautiful and old-fashioned kind of "joined-up" script, I was both amused and envious.

  28. Tom O'Brien said,

    April 1, 2014 @ 12:01 pm

    I'm happy to see that my patron saint's handwriting was no better than my own. If the nuns had known this, they might have cut me some slack.

  29. Levantine said,

    April 1, 2014 @ 1:18 pm

    Sorry, I meant back in c. 1990 in my earlier comment!

  30. Wendy said,

    April 4, 2014 @ 11:56 am

    @Levantine–And now I know what Gilderoy Lockhart was so proud of learning! Thanks! :)

  31. Geoff said,

    April 6, 2014 @ 6:43 pm

    My mother's writing (born 1926) is beautiful, but to me it's just a string of m's and u's.

    My writing, I modestly claim, is a reasonably neat italic (fashionable in 1960s Australia)*, but I'm sometimes surprised that people a generation younger have trouble reading it.

    * We still used ink pens with a nib designed to create the thick and thin strokes. A pain in the neck for a left handed 7 year old, of course.

RSS feed for comments on this post