The Development of Writing and the Progress of Enterprise Software

futurelab default header

by: Sigurd Rinde

The simplest concepts often take the longest to develop.

Writing, with hindsight, seems to be a stupendous simple concept, still it came into being very, very late in the history of mankind going through three distinct phases (so far): From very complicated and inflexible towards relatively simple and thoroughly flexible.

No doubt that the first practitioners found the power of writing extremely useful and defended their own exclusive use with all means – the most efficient one being "keep it complicated"! Thus the development was reined in, delaying the progress even more.

Keep this thought as you ponder the huge and complicated IT systems running big business today; could there be other reasons than technological for the lack of quantum leaps in development?

But first, how would I describe the concept of writing?

A representation of sounds or concepts and their sequence so spoken words or facts can be recorded, distributed and recreated.

And how would I describe the concept of enterprise software?

A representation of things or concepts and the sequence of things happening so activities can be recorded, distributed and recreated.

In it’s most obvious form this would apply to process oriented software as is the norm for many Enterprise Systems. But in truth it does apply to all kinds of software as well, from photo manipulation to social networks – as every activity is a process, a sequence of things happening to be recorded, distributed and recreated.

Lets have a closer look at how writing developed, then compare it to the progress of enterprise software:

1. One to one representationLogographies

The first natural step towards a writing system – pictures, logos, one symbol representing a full word or even sentence. Later this developed into mashups of two or more "logos" – say "house" + "man" = "home".

Simple as concept but soon becoming quite complicated and rather bloated, and definitely not very flexible – not much humour or poetry expressed, mostly accounting for real-world objects. Man owes four oxen and one pig type of prose.

Some have developed further and are still in use, Chinese characters comes to mind – beautiful to behold but it requires a Chinese literate to remember about 4,000 symbols.

2. Few to many Syllabaries

The natural next step, the first effort to atomise the words by representation of sounds you hear clearly; syllables. With Syllabaries the power of recreation increased, the number of signs to remember decreased. Still rather complex, so not much humour nor love stories yet.

3. Fewest to everythingAlphabet

Eventually, with a quantum leap in inventiveness, a singular representation of the smallest real-world communication object; the single sound. Then relate each letter in linear sequence to form words that again relate to each other to form meanings.

The number of symbols decreased again and the number of words, sentences and meanings became limitless allowing for it to spread beyond the palace based scribes. At last humour and poetry came to life. And with popular use some of the power structures came under pressure, in fact the first alphabets and democracies happened at more or less same time and place.

Still not perfect though; it’s lacking the ability to capture, distribute and recreate directly and simply the time context. That still requires many paragraphs in the right sequence and representation of many real world objects (both concrete and abstract), all in one place. Thus the media expanded from simple tablets, drinking cups and plates to scrolls and later books, documents and forms. Enter historic source analysis, misunderstanding, misrepresentation and, as is highly visible these days; twisted political ads in heated Presidential elections.

Now to the information technologies. Obviously IT started where writing left off, but there are progress parallels:

1. One to one representation – Documents, Forms and Transactions

Starting with the latest writing technology and representation methods; nothing more elaborate than representing the documents, forms and transactions as if they’re first class objects was conceived. And that in simple terms is where we are. More or less.

Even an effort to move away from paper-forms in bits and bytes, replacing them with IT specific objects could not free itself from the old ways: Ask SAP what "BUS2089" is and you’ll get the answer "Employee Trip Business Object". A single multi-object representation for a distinct abstract notion. Very Chinese.


With this model the number of "information representations" simply explodes, Chinese becomes real easy in comparison. It’s usually called "information overflow" and the unscalable wall of complexity looms.

2. Few to many – Tags, Search, Non-double-entry book keeping

Now this is somewhat different from the Syllabaries, but still represents the next logical step forward.
The forms, documents and "business objects" still exist but tags and other metadata has been added so each representation makes more meaning. There’s a clear drift towards this lately, and exemplified by Workday who even moved away from the pure context based transaction representation (double-entry book keeping methods).

3. Fewest to everything – Singular objects, Semantic relations, Activity flows

At last the documents and forms can be split into the smallest singular objects using relations and records of any change to allow for no limits to representation and recreation, the lowest possible complexity (exactly same as reality) and far fever symbols/functions/representations to remember. The technology exists, the mind lags.

As you would expect, this is what Thingamy is trying to do – introduce the software analogue to the alphabet with (currently) 29 "symbols" to represent any real world object, activity and time context.

But, "no model nor theory is ‘right’, there are only some that awaits being proven wrong". To get anywhere one have to work hard to prove the current "right" model wrong, then start the work to prove the next model wrong, rinse and repeat. No time for rest 🙂

Original Post: