Death of Writing, pt1: Organization of Information

Send to Kindle

Hmm, what else can we pronounce dead? Oh yes, writing!

Writing is dead.

(he writes)

Writing, as in human authored, textual representation of human language.

Dead, as in soon there will be very smart and successful people who will not know how to write.

Dead, as in we will gain much new knowledge without the use of writing.

(Not dead as in totally disappeared, but dead as horses are dead as the primary mode of locomotion. Or film cameras are dead as the primary way to take photographs.)

{Sure, information will always be encoded in some way, and for our convenience it may be presented as text. But when information doesn’t represent human language, it is not writing.}

Act I, Scene I.


Written resources are worse than my phone at helping me find a restaurant I will like, therefore writing will become obsolete.

Writing is a great technology for creating resources to be used by many people spread across time and space. Writing is well suited for providing universal answers, for creating resources that have to be everything to everyone—such as encyclopedias, dictionaries, or textbooks. But such universal answers are no longer sufficient for me (I don’t want a listing of all the restaurants in the whole city, Yellowbook.) I want a particular answer to a question I have here and now (I want the restaurant that suits my taste, that is near me, and is open right now, iPhone Hi Galaxy.)

Writing cannot deliver those kinds of solutions.

My pocket device can, however.* It can give me just the nearby restaurant I will love. It can do so, and this is the important bit, without relying on writing.

{We are moving from universalizing to particularizing}


Think about the organization of information about restaurants: It is impossible to create a written resource that lists restaurants in actually useful categories like “falafels near 15th and Walnut,”  “restaurants that Joe and Keisha will enjoy on their first date,” or “just the trendy restaurants Sylwester will want to visit on February 15th.” Those are categories that correspond to questions that people have.

Instead, when writing still meant printing we created categories that serviced no particular queries and then we taught users how to obtain from those categories answers to their particular questions. When you think about it, it’s a roundabout way of getting at things, necessitated only by the limitations of writing—it’s inability to change and to respond to the user. We grouped restaurants by cuisine type, by (someone else’s) ranking, by neighborhood. Both makers and users learned this system (to the point that this kind of organization now seems somehow natural and is still with us in resources like Yelp, Zagat, Urbanspoon which are only starting to move beyond their print predecessors).

Rather than a system I have to learn and that provides only somewhat useful knowledge, I want the information system to learn about me and provide for me exactly what I want (“this is a restaurant you will enjoy right now” is more useful to me than “three star, moderately priced Afghani eatery in my neighborhood”).

Again, writing cannot provide such answers.

Things that are written down are dead. That is, they cannot change. They cannot respond. They do not know who I am.

Written resources have to anticipate my questions and pre-generate their answers. That limits their usefulness.

The written text conveys something as it was for the particular author trying to anticipate who I will be and what I will want; trying to guess her audience and their needs.

Written text cannot respond to questions (what do you think of the Indian place down the block?). It cannot respond to feedback (I liked the saffron rice you dissed, how does that change your recommendation?). It cannot adapt to new information (maybe for the reviewer this was the best Afghani in town, but a new one opened since, what about it?).

My phone relies on algorithms acting on big, context specific data to provide answers for me. Those algorithms do not have to pre-generate answers, they create them when needed. They don’t have to guess who I am, they know all about me, my friends, my enemies, and strangers that are just like me. The algorithms are never finished with their work. They provide solutions and then await my response. A Google search algorithm awaits which of the links I click and then prepares itself to be more useful next time. Writing has one answer and it cannot change.

Most importantly—this is the crux of the argument—my pocket device provides me the answer without analyzing writing. In fact, it could not get the answer by analyzing writing because the answer is based on information that is never written down. Instead, it is a digital analysis of actions: where did I go (online and in real life), whom did I friend on my social network, which like buttons did I push, which like buttons did my friends push, and such. None of those things are textual representations of human language; they are digital traces of my actions. This is the kind of information that is used to provide me a more useful answer to my query “what restaurant do I want to go to?” Writing has no chance to compete.



Writing is not very useful for organizing information. (But you already know that, because you would rather ask someone you trust, than read something written by a stranger.)

Writing is being replaced as the primary technology of knowledge organization by algorithms acting on big data.


Writing is dead (he writes)

To be continued.

*If not exactly now, soon enough.

Send to Kindle
This entry was posted in classification, death of writing. Bookmark the permalink.

2 Responses to Death of Writing, pt1: Organization of Information

  1. Pingback: Writing is Dead, pt 2: Interface » 23. {insert footnote}

  2. Pingback: Death of Writing, pt 3: Generalizations » 23. {insert footnote}

Comments are closed.