You are on page 1of 7

On Logographic Characters

The goal of this paper is to summarise some preliminary research and to introduce some ideas
concerning the development, use and possible future applications of logographic writing systems,
namely the Kanji, or Chinese characters.
This ancient system of symbols may not only hold interesting clues about how human cognition and
non-verbal (pictorial) communication may have evolved and be currently working still, but it may
also pave the way for a new method of IT programming which could potentially lead to a
computational language far superior in speed, coherency and efficiency than the binary linear one in
use today.
As a disclaimer, I would like to state that I am not a native of either language using such characters,
I am simply a student of Japanese who has delighted in seeing interesting similarities between her
own background in Physics and the language she has committed herself to learning as completely as
possible.

1 Kanji as natural science archaeological artefacts


According to Japanese folklore [1], a government official of the legendary Yellow Emperor got the
idea of making the predecessor of hanzi/kanji from looking at the markings and footprint of
animals, very much like the fanciful 'dinosaurian' letters of the imaginary world of James Gurney
(Dinotopia, [2]). Uncertain as this origin may be, these characters most certainly have their roots in
the mimicking of Nature and the surroundings of humankind.

Source: Wikipedia (Dinotopia)

As such, they provide us with unrivalled insight into what our ancestors of that area understood of
the Natural world, their level of science and cognitive abilities. Literally, they have drawn us
pictures of what truths they saw, or professed to see in the world around them.
The oldest surviving piece of such writing that we know of today comes from China, the so-called
oracle bones [3]. In Japanese, the characters found on these pieces of shell known as the koukotsumoji. The literal translation may be thought of as the kanji version of the English pun; upon first
hearing, knowing the context for the new word, one may think that this is written as , that
is 'hard-bone characters'. However, the official written form of this word is the phonetically
identical , which gives the literal meaning of 'first-bone characters', a delightful play on
kanji use and one that seems very typical of the language. Association between ideas can often be
done phonetically, simply because any sound may be attached to a character, even two or three very
different looking and meaning ones, thus arguably enabling the language to keep a relation between

them based upon the using community's world-view without compromising the precise meaning.
Thus, the spoken language is fluid and initially may seem indefinite and strongly based upon
context, however the written language is precise and open to new levels of meaning that a nonvisually based, phonetically spelt language is simply incapable of showing, regardless of how many
composite words and conjugations it may use. A picture may thus truly tell a story of a thousand
words, as the saying goes.
Another very interesting and not to be missed importance of the use of hanzi/kanji is its unifying
nature. The very fact that the shape of the given written information has no set pronunciation, unlike
phonetic languages, also means that people may still talk in their own native languages while still
being able to completely understand one another. [1] Each group of people may give their own,
relatively arbitrary sounds to the written form as they please while still keeping the same meaning.
Two people from the two far ends of the Earth, or, even the very Universe itself may meet one day,
have no idea about what language the other speaks and still be able to communicate precisely and
intelligently with one another through such a writing system. There would be next to no need for
one language to overtake another because all would be equal in terms of written communication.
Native languages would thus be able to survive in their given localities while still forming
important connections between those localities, forming trade, alliances and unifying even vast
areas with a plethora of different peoples. Variants of the writing system would, of course, arise in
time, however even today people from China and Japan, each having very different cultures and
languages are fully able to communicate with one another simply through this highly ingenious
writing system.
And thus the idea arises if it would be possible to create a similar or variant of hanzi/kanji which
could be used as a world-wide writing system, thereby rendering all native languages equally
important. If so, then most or all languages would be able to be kept from extinction, tensions
between regions might possibly lessen in terms of understanding and, even, new avenues of trading
and other multicultural exchanges might open up in and between previously isolated areas. At the
very least, the rate of information exchange, the driving power behind the modern world's 'stock
exchange' would significantly increase as the much-feared 'language barrier' would greatly
weaken.
Now, if we look at the following kanji groupings and , as well as, and , we can see that
in each case there we have an initial character from which the second is obviously derived through
the incorporation of the first.
In the first case, the derivation is quite simple, even pictorial. The character stands for 'tree' as a
pictorial representation in itself, with branches originating from a trunk, while very simply and
concisely represents the idea of 'forest' as a concept which incorporates many 'trees'.
In the second pairing, we have as the character for 'water'. The derived character, , is
comprised of two components (i.e. radicals), namely one for 'point' and the other being the previous
'water' character, stands for the concept of 'eternity'. How this this concept-derivation made? One
must simply do what the old folk-tale of what the Yellow Emperor's government official is reported
to have done and turn to Nature and observe.
If one looks at any point in water, one cannot find any starting point, any real 'end' to the material.
Definitely, even if one takes the water out of one container, it can change into a seemingly infinite
amount of shapes, according to the container. Even just watching the water-cycle as we are taught
about in science lessons about ecology and the environment, it most definitely seems the very
definition of endless; very much the 'chicken and egg' type of concept of infinity. Thus, taking any
'point' of 'water' gives us a very easily understood explanation of something as heinously difficult to

understand as 'infinity'. And, delightfully enough, not just one 'type' of 'infinity' is given, but
immediately the three above: spatial, structural, and temporal, making concept conveyance that
much more layered and complete. Instead of the long explanation written above, all one has to do is
read/recognise the kanji and all the information above is immediately known.
Again, the possibility for gaining insight not only into the language, but also the observation-based
philosophy and science of a long-past era of human thinking that is still valid today must be made
clear here.
As can be seen from the above, hanzi/kanji have, through a mix of design and what could be called
'memetic-evolution' of concepts become a system based upon the natural recognition abilities of the
human brain thus enabling meaning to be perpetuated simply using basic knowledge gleaned from
a common, surrounding natural world every human being has access to, experiences and observes
on a daily basis unconsciously, if nothing else, making the both the basic and derived characters
'obvious'.
[1]

5/6/7

ISBN978-4-89096-102-3 c0081
[2]
https://en.wikipedia.org/wiki/Dinotopia
[3]
Oracle Bone Cambrige University Library
http://www.lib.cam.ac.uk/mulu/oracle.html

2 Kanji and Facial recognition


Human beings have evolved towards reading and understanding the faces of their fellows. As such,
it is of no real surprise that this habit has become something of a talent for those who have managed
to survive the long-haul of the natural process of selection. Body language and facial expression are
so important that even using 'faceless' social media, people find a way of inserting pictorial
representations into their messages (emoticons, memes, gifs, etc) as a way of giving context that
seems to be missing and might easily be (and often is) misinterpreted without the ability to read the
face on the other side of the line.
Upon meeting kanji for the first time, the sheer amount of information in strokes and complex
shapes can be overwhelming. They can appear quite alien in the eyes of the non-native or untrained.
However, the learner will soon recognise that there is method in the apparent madness of those
agglomerated abstract figures. Not only is there a stroke order to be strictly followed (from upperleft to bottom-right) to be able to more easily recall and write a recognisable figure, but also the
groupings of the radicals into sections of the given square area they define is also of utmost
importance.
The question then arises: why has such strict pattern of memorising and replicating these condensed
information blocks develop in this particular way?
While there seems to be a general consensus that an average language-fluent person uses about 2,

500 kanji in everyday life, a more-highly read or educated person will naturally know more. For
example, someone who as passed the Kanji Kentei's () Grade 1 would, in theory, be able
to read around 6, 355 kanji. What seems to be the upper-most said limit are those Japanese people
popularly called Living National Treasures () [1] who are reported to be able to recognise
around 10, 000 characters.
This number is familiar in that this is the estimated upper limit of faces a human mind seems to be
capable of recognising [2].
However, though this could simply be coincidence, it does raise the idea of there might being some
link as to how kanji stroke order has developed in the way it has and how it is natural for our eyes to
move when we are trying to recognise a given face or pattern.
There have been a great many researches have been done not only to track the movements of our
eyes when we recognise someone, but recently new insight is slowly being accumulated as to what
the neurological and psychological reasons behind this activity may be as well as what connections
may be betweens the two.
The similarities between kanji and a human face cannot go unrecognised. Both comprise of 'blocks'
of information, each of which has its own identity and build together to form a unique and uniquely
identifiable whole and these are each 'piece by piece' run over by the reader's eye in order to be able
to identify the whole. The radicals of a kanji may be compared to the distinct features on a human
face, i.e. the nose, mouth, eyes, eyebrows, chin, etc. Just as the kanaka may be divided into areas, so
can the face (i.e. Upper-left, bottom-right, centre, outer, etc.).
In such research, the points at which the participants' eyes focus on and collect information for
recognition is called the fixpoint and the saccade is what brings the eye movement to the next
fixpoint. Based on research ([3], [4] and [5] ) the place where an observer's eyes stray to during the
process of recognition have real, functional roles and are far from being chosen at random.
In an article by Norton and Stark [6], the notion of a so-call scan path theory was introduced and
has remain influential, though still a subject of debate, today. It connects both visual perception as
well as the related motor functions of the eye to being induced via the inner cognitive
conceptualisation happening with the human brain. The idea is that there is a fixed pattern of
fixations and saccades which an observer's eyes will follow during the recognition process.
Though various research seems to have the success rate of this idea to be at only about 65% or so,
could a trained scan path be used successfully as a mnemonic device to aid with the visual
recognition and memorisation of thousands upon thousands of characters?
Interestingly enough, it has been suggested that the very nature of what is being observed does have
a significant influence upon how likely a given, individual scan path will be replicated [7].
Seemingly proving this idea, when subjects are given a pictures to memorise then are asked to recall
the picture staring at a whiteboard, those who were asked to only look at the centre of the board
scored a significantly lower success rate than those allowed to move their eyes within the board
parameters [8]. Another interesting result of this research was that the scan paths of those subject
who were allowed to have their eyes wander closely matched the movements their eyes has made
previously while looking and memorising the picture, suggesting that such given scan paths may
indeed be used as mnemonic devices in aiding the storage and recollection of pictorial information.
In a recent article [9], researchers found using Asian participants that the average preferred scan

paths started from the right eye, then moved towards the nose tip, remaining in or shifting between
these regions. There was only a low possibility of the scan path starting from the left eye, but in the
cases it did, it tended to either remain in this region or move on to the right eye. Also found was that
correct and incorrect recognitions could be linked to distinctive patterns of the eye, with the
emphasis being on the transitions between the fixations instead of only the points of fixations
themselves.
Questions arise on whether or not recent studies into the neurological side of the cognitive
experience could shed light onto how and why hanzi/kanji have taken the shape that they have, and
have their stroke order defined as it has been.
Also of considerable interest would be why this seems to have developed into a naturally defined
optimal specification for such information exchange. Understanding this could potentially further
our understanding of how recognition works, of the conceptualisation process itself and how nonverbal information exchange may work between members of the human race.
Yet another application of this idea may be that if we are to have more precise machines and
programs capable of mimicking human thought, then one would argue that having these follow the
mechanisms for human thought would be a prerequisite for its proper function. Certainly, it would
be necessary for users of this technology in order to be able to easily understand and interact with
them.
[1] https://en.m.wikipedia.org/wiki/Living_National_Treasures_of_Japan
[2] https://www.theguardian.com/uk/2004/dec/13/sciencenews.research
[3] Henderson, Williams and Falk (2005)
In this research, when participants were restricted to viewing only the central area of the face (i.e.
nose, eyes, mouth), their performance rates in recognition showed a significant drop.
[4] Rayner (1979)
The preferred viewing location for readers of English is between the beginning and centre of the
given words.
[5] Henderson (1993)
The preferred viewing location for observers of objects is the centre.
[6] Norton & Stark (1971)
[7] Josephson & Holmes (2002)
[8] Laeng & Teodorescu (2002)
[9] Tim Chuck, Anton B. Chan, Janet H. Hisao (September, 2014); Understanding eye movements
in face recognition using hidden Markov models.

3 Kanji and It applications


In terms of Information Technology, the applications of understanding how kanji truly work would,
perhaps, have the most far-reaching applications. The idea of not having a linear basis for
information storage and processing like the binary system, but a 2-dimensional language for modern
technology to 'think' in would greatly increase not only the speed of processing and the storage

capabilities of systems of the same size as contemporary ones, but would also mean greater, or, to
be more precise, more human (if one may say so) thinking from our machines if we should choose
to give this to them.
Denso Wave [1] has developed a kanji-based coding system, the QR system which was originally
developed for the systematising of its own products, but has gained world-wide favour and use as
being a simple and effective way of utilising existing cellphone technology to link the average user
to websites connected to products they might be interested in or, even find profiles of friends and
acquaintances on social networks without having the go through the tedious task of inputting long
URLs or confusing and obscure usernames. Just one simple snap using the mobiles device's camera
of the QR patch and the user will be directly connected to the page of choice.
Compared to the linear barcode coding, the QR code exhibits many points in which it has proven to
be superior in use and function. Their being able to be scanned far faster [2] and more reliably [3]
store greater amounts of data in a smaller area [1] than the traditional UPC barcodes, not to mention
be 'readable' from any 360 direction has meant that they have been able to spread far beyond its
humble beginning as as a method for helping logistics use within the automotive industry.
This idea applies not only to data storage, but also to coding. The idea of having a programming
language which is not based upon a linear string of zeros and ones, but upon blocks of 2dimensional would naturally produce the same results of greater speed, data amount, and reliability.
However, what might be an even greater possible consequence arising from the use of such
programming could well be true pattern recognition from the machine.
It is a long-standing complaint from anyone and everyone who has ever used a computer or related
device that it is a 'stupid' machine. The reason behind this statement is that a computer does not
think like a person, it does not recognise patterns and connections that a human does, it cannot
understand what a human would consider to be 'important' because its logic, while completely valid
in terms of the science of Logic, is not the insightful and what could be called 'intuitive' ability of a
human.
As an example let us say that in order to make, a bowl of soup one would need to have the
instruments needed (a pot, a stove, a bowl and spoon) the ingredients required in given amounts
(water, onions, carrots, meat, oil, salt and pepper) and steps which one must take in a given order to
attain a fresh and delicious bowl of soup ready for human consumption at the end.
With these, both the average human and a task-programmed computer would be able to complete
this task, but if anything were to change the human would be able to adapt, while the computer
would not. For example, if the task were to start not from scratch, but from the water already
boiling. The human would simply continue on from there while the computer might either continue
as if the pot were empty, completely ruining the meal, or empty everything and starting over,
wasting both time and resources. This is what a person would call 'stupid' behaviour even though
the machine is following a completely logical line of reactions. Of course one might program the
machine to recognise this, but then what if instead of carrots, the machine has potatoes? This,
naturally would not cause the human any problems, but the computer would not even be able to
start the meal because there are no carrots, what it considers to be a vital part of the recipe. 'Stupid'
behaviour once again, as one knows that carrots may be replaced with potatoes as these are also
edible, but the computer has nothing to base this on. Again the program can be expanded, adding
'edible foods' that any missing ingredients may be replaced with, but another problem immediately
arises. Chocolate is 'edible', however not something one would normally want in a meat soup. One
would have to go into even more detail in programming to account for the huge reservoirs of human

experience and create large, bulky and cumbersome databases for even the simplest of human
endeavours.
And the problem still remains: reality always comes up with a new situation to be negotiated which
the programmer might not have thought of if for no other reason than not having yet experienced it
himself or herself, while the physical size of computer storage is limited and a packed memory
considerably slows processing, regardless of advances in hardware technology. I suggest that this
problem is not only to be solved through advances in hardware alone. After all, what use is having a
much-coveted quantum-computer if you can only run the same programming on it that Babbage's
Difference Machine is based upon? It would be somewhat equivalent to having a Ferrari chassis,
but only putting a lawn-movers motor into it.
Now, what if databases were based not upon linear streams of numbers (which also, it must be
mentioned, has a considerable translation error as any analogue-to-digital conversion has), but upon
a system of interconnected nexuses of related data stored in 2- or even 3-dimensional blocks
between which visual-pattern recognition programs can easily run through to create connections
between what is observed by the machine in the surroundings and what it has now far more
conveniently and accessibly stored in its memory. To go even further, if the processing of the
machine is also based upon working with a 2-dimensional, that is a pictorial language of far greater
nuance capabilities than a linear language could ever have, then the rate and precision of these
connections would greatly increase.
This would also have the result of the machine being able to gather, process and apply new
information with far greater efficiency and, perhaps, help pave the path for the first truly selflearning-capable computer.
[1]
Denso Wave hompage
http://www.qrcode.com/en/about/
http://www.denso-wave.com/en/adcd/qrcode_20th_anniversary/
[2]
"QR Code Essentials". Denso ADC. 2011. Retrieved12 March 2013.

[3]
https://en.wikipedia.org/wiki/QR_code#Error_correction
Summary
This essay has no goal other than to suggest ideas that I have tried to superficially look into over the
course of the last year, as allowed by my own time, language knowledge and background in
Physics. I would be delighted to have any help possible in following and why I believe these areas
of research could prove not only beneficial, but uniquely so for a number of areas of both Science,
Technology and Philosophy, not to mention for people in general. As can be imagined, these are all
primary reasons why I am determinedly pursuing my studies in Japanese and hope to be able to
make a successful N1 exam within three years. By that time I hope that these ideas might become
better streamlined, detailed and I will have better capacity and ability to follow through with
developing either one or some or all of them as far as it is possible.

You might also like