Considering the variety of systems of writing, the ease with which someone can receive written information in one system of writing is not precisely identical to that of any other, and I am curious to learn from any existing efforts to analyze such information. I would also like to review others' efforts at analyzing the ease of writing compared to that of reading for any given language.

Of course countless individual and social human factors confound the possibility of meaningfully quantifying and comparing such characteristics of natural languages' reading and writing systems. Nonetheless, I trust that I am not alone being obsessively interested in the implications and effectiveness of humanity's efficiency interacting with methods of language input and output.

What disciplines are currently engaged in analyzing perspectives related to efficiency of writing systems, and when did they come into existence?

3 Answers 3


I do not agree with your assessment of the facility of reading Chinese: I find it extremely difficult to read Chinese compared to English or even Russian. So there are probably other factor at play, such as the native language of the subject. Is there some experimental literature that supports your belief?

Theoretically speaking, this would be a question to be researched within the sub-discipline of psycholinguistics, which in one view came into existence in the 19th century. The problem which plagues questions about the efficiency of writing systems is the problem that plagues all questions about efficiency of human actions, namely the lack of a good scientific framework for measuring efficiency.

  • 1
    The problem with the whole question is that there is no consistency between individual people in how they read. Humans have no genes or adaptation for reading the way we do for talking. Reading is technology, and everybody learns to cope with technology on their own terms, using whatever resources they can muster. This results in vast degrees of variation among individuals with a common language; the cross-product of this variation with the variations in actual orthographies is really vast variation from individual to individual. So vast that it's unmeasurable in any real sense. Feb 16, 2015 at 18:47
  • @john lawler in exile: I know software can't yet "understand" natural language, but can't the ability to convert between formats (text-to-speech, and to a more limited extent, automated translation between languages) be treated as a proxy here? Are some written languages known to be easier or more difficult for software to handle in this context? Feb 19, 2015 at 17:09
  • Nothing is known about most languages because nothing has been done with them. And the ability to convert text into speech has nothing to do with translation -- it's strictly sound production and dictionary lookup; speech to text is much harder, because speech is not atomic like text. And automatic translation is ... as you see it -- still being given away free, still definitely a beta enterprise, not ready for prime time. Feb 19, 2015 at 17:32
  • Google translation and synthesis for Norwegian is vastly better than it is for Swahili. This is not because Norwegian is "easier", it is because of the massive asymmetry in knowledge and computational energy available in the two languages.
    – user6726
    Feb 19, 2015 at 17:54

The ease of perceiving a written text is, indeed, an unmeasurable thing, since it take too many factors: sociocultural, historical, and even medical.
Also, note that there is, in general, many-to-many relation between languages and written systems. For instance, Azerbaijani, Romanian, Serbian, and Uzbek are/were written in Cyrillic or Latin.

On the other hand, you can measure (and improve) the effectiveness of understanding a written text. There is a science studying this, it is called Rapid Reading. This technique is intended to increase the amount of written text per unit of time, so, naturally, it can be measured.

There are many factors slowing down the amount of text we can read and understand per unit of time. These factors are:

  1. Sub-vocalization, a phenomenon when a reader reproduces sounds of the text;
  2. Seeing a fraction of the page at single moment;
  3. Linear reading and "rolling back" to a different place in a previous line to re-read;
  4. Reading every letter in a word is not effective, since we can normally understand words wiht minor typos. This suggests there can be a more effective writing system.

Some writing systems are free of some of these negative factors. So, an educated reader (not a language learner!) may benefit from those. The most evident example is Chinese. Consider:

  1. Since the written word has (almost) no relation to how the word sounds, a reader would not sub-vocalize it;
  2. Writing systems that use blocks (not linear words) can be also read more effectively.

This answer at Chinese.SE provides with some more details on the topic.

  • You mentionned the fact that we could improve efficiency of written text (it's information density) by removing redundant features. But let's keep in mind redundancy does matter in order to limit errors. Think of the NATO alphabet : it uses whole words (up to 3 syllables !) to code for 1 letter, because it has to be recognisable in a jammed radio conversation
    – Axel B
    Mar 26, 2020 at 14:15

The tenth chapter of Geoffrey Sampson's Writing Systems: A Linguistic Introduction (1985) discusses English orthography, and by and large reaches a conclusion which somewhat resonates with yours; that logographic writing (such as Chinese) is efficient for the reader, somewhat at the expense of the writer, and that the quirks of English orthography push it in the direction of obtaining some benefits of a somewhat more logographic approach (so the different and sometimes decorative spellings of some homonym pairs in English accidentally but conveniently help the reader quickly distinguish between them).

Sampson quotes many sources, some linguistic (e.g. Chomsky & Halle 1968) but many others from "the psychology of writing". I don't know if this is a proper subfield of psychology, or just Sampson's informal label for these studies. Notable references include Albrow 1972, P.T.Smith 1980, Wijk 1969, Frith 1979, 1980, and Bryant & Bradley 1980.To the extent that Sampson refers to (some of) these authors collectively, they are called psycholinguists.

I'll also not that the Wikipedia article on Psycholinguistics has a separate Reading subsection though it mainly discusses the basic motorics around reading.

  • I see there is now a revised and expanded edition from 2015. For the time being, I only have access to the original edition. Let me know if you want these citations in more formal or detailed form.
    – tripleee
    Mar 19, 2018 at 19:20
  • The question was substantially edited so "your conclusion" is only visible in the edit history now.
    – tripleee
    Jun 25, 2018 at 12:21
  • Reading the second edition now, the discussion about reader vs writer efficiency is a recurring topic in the book, where e.g. the discussions around Arabic, Hebrew, Greek, and Chinese script all touch on this subject, as well as the related topic of deep vs shallow orthography and its effects for readability. Subjectively, I find languages with shallow orthography (Finnish, Norwegian, German) noisier to read than e.g. Swedish and English.
    – tripleee
    Apr 18, 2020 at 8:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.