This could be one of the greatest unanswered showerthoughts of all time, and unsurprisingly many had raised the same question on Yahoo, Reddit, and Quora (as well as its Chinese counterpart – Zhihu), although the majority of them stuck at the level of ‘Speaking Cantonese grants you 15% off at your local Chinese takeaway’. Admittedly it is kinda a big deal, but what we as amateur designers and trainee psychologists are really interested in is ‘Does knowing Chinese enables one to think in an advantageous way?’.
In what way could a language be advantageous?
It goes without saying that the written characters (glyphs versus ideographs) vary dramatically from English to Chinese; they are undoubtfully the most diverse languages among the major languages in the world. Such a difference goes beyond the structure and appearance of their characters, in fact, the most differentiated part in the comparison is the information they are carrying. The concept of entropy is introduced here to measure the information content encoded in a natural language; it calculates the redundancy and repetition in the language based on the uncertainties when forming sentences (Shannon, 2001). According to which, An example of an ideal data compression should satisfy that:
a string with entropy H(X) can be compressed into more than N H(X) bits with an insignificant loss of information
I am not a big fan of numbers, but all we need to know is that the concept of entropy established by Shannon made it possible to scientifically compare the degree of preciseness (as less redundant languages being more precise) between different languages. Unsurprisingly, researchers found that Chinese, among other mainstream languages, in both modern (United Nations Charter) and ancient (Holy Bible) literature, has text length of the smallest size. Moreover, even when taking a less systematic approach through comparing average texts after translation, we will still arrive at the almost identical conclusion, as The Economist had shown in one of their daily charts.
Smallest in size in both conditions (Behr, Fossum, Mitzenmacher, & Xiao, 2003)
The gain or loss in text length after lossless translation (The Economist, 2012)
The magic behind the efficient encoding of Chinese language, as we all know, is the power of ideographical scripts. Which enables a simple character to convey a message that is otherwise requiring a full sentence to describe. This leads to yet another fundamental difference between Chinese (and other East Asian languages share the Chinese characters) and languages use Latin alphabets – Phonemic orthography. While one can attempt to pronounce a strange English word, the same could seldom be done with Chinese because the information for its pronunciation is embedded in the Chinese characters. Chinese speakers receive very limited cues when encountering an unfamiliar ideograph, which explains the relatively high illiteracy rate in China as structured training is required to associate the words with their meanings and pronunciations.
Does our brain know the difference?
It is truly amazing how our brain could adapt these differences when guiding us to acquire language ability, it now leads us back to the title; Does it grant you any advantages in the cognitive process for knowing Chinese? First, it is crucial to address if there really is something going on in the brain of Chinese speakers, as in respect to the Latin-based language speakers. Naturally, the first stop of the investigation are the classic go-to areas when it comes to language production & reception – Broca’s and Wernicke’s area, located in the lateral frontal lobe and superior posterior temporal lobe respectively.
The go-to places when it comes to language processing
Speak in different tunes
Through fMRI causal modelling, activation of both areas was found in speakers of two groups when generating and perceiving language. This is hardly a surprise given how fundamental the two regions are in the language processing. However, when moving on to the connection hub of these two areas – anterior superior temporal gyrus, it is revealed a much stronger connection between this hub and two language areas among the Chinese speakers. It is widely accepted now that the increased connectivity is a direct result of refined ability to handle tones in the language, which is one of the unique features of Chinese. Furthermore, it is also discovered that a part of RIGHT hemisphere – the right superior temporal pole, could also be responsible for the similar function, breaking the made-believed dominance of left hemisphere when it comes to language processing (Jianqiao Ge et al., 2015). It is noticeable that the introduction of this bilateral language network is mainly attributed to the fact that Chinese is a tonal language with an adaptive demand of enhanced mapping of sound. In other words, Chinese speakers have a superior ability of speech hearing, although the evidence beyond language processing is still limited.
Result of the Dynamic Causal Modelling (Ge et al., 2015)
Write in different scripts
So, how about the written form of Chinese? It is cool to see a remarkable revision on our language network, perhaps we will see musical academy teaching Chinese just to enhance students’ ability to catch the perfect pitch. But what are the cognitive benefits of knowing the representation and pronunciation behind each squarish boxes? By the end of the day, that is the single most important feature to the designers.
Once again, the cognitive neuropsychology approach saves the day. The learning disorder known as dyslexia is a reading difficulty which prevents one from the normal efficiency of spelling and reading, often in a way that words in a sentence could not be effectively recognised because the patient has trouble 1) integrating a word through its components 2) associating the pronunciation. Dyslexia is commonly regarded as a disorder specific to users of phonemic orthographies, such as the abovementioned English, Spanish, Turkish, or even Vietnamese after romanisation. Intuitionally, children from east Asian countries, where languages have their pronunciation embedded in each character, wouldn’t show the same disorder (it wouldn’t be a difficulty if it’s not even deemed as a function). There is, however, another variant of dyslexia goes by the same name but a slightly different mechanism. East Asian children with dyslexia have trouble telling a non-existing ideograph from the real ones, and noticeably slower when identifying the meaning of a given character. Study has shown that there is cognitive difference between the two disorders, while children of both groups show activities in the left temporoparietal cortex, children from China exclusively showed decreased activity in their left middle frontal gyrus, which had been confidently associated with shape coordination and recognition (Siok, Perfetti, Jin, & Tan, 2004).
For different orthographies, dyslexia takes place at RELATIVELY different stages of recognition.
Put a smart filter onto your text layers
And now let’s put all these distinctions of Chinese back to the world of machines, and see if there is something we could do to optimise the experience of users from East Asian countries. In HCI 101, we learnt that language as a representation of output is mainly considered as an attentive information, and occasionally, passing through the pre-attentive channel. The attentive information is processed by the prefrontal cortex as several types of executive functions (such as decision makings and outcome evaluations). The most intriguing part, however, is the pre-attentive feature of language in the interface, as which it influences the subconscious thoughts of users and hence guided users to overt behaviours. An ideal pre-attentive stimulus should make users’ pattern predictable and desirable, which happens to be the exact speciality of language.
This flow chart of mental model sums up the interactions between gyrus and chips
Based on the neural mechanism of dyslexia, it is most certain that ideographs are one step sooner than alphabets/words when matching with meaning in the reader’s mind. Chinese characters can be associated with corresponding meaning without the realisation of phonological awareness, which as a direct result makes its speakers less vulnerable to dyslexia. In other words, a Chinese word on the screen can be understood faster than its English equivalent. I could think of two underlying explanations for this phenomenon other than the fact that phonological process happens simultaneously with orthographical matching. First, the shape recognition process of a Chinese speaker, just like sound mapping, is largely enhanced through years of training. Second, when breaking down the Chinese characters, they almost always consist of subordinary characters of what we call radicals. Before recognising a character, it is essential to identify each of its radicals, doing which is already a process of acknowledging the meaning of these radicals on a subconscious level.
The second explanation – a ‘latent recognition’ parallel to the standard process
The straightforward implication here is to simply adopt the Chinese character in icon and logo designs so as to speed up the pre-attentive process. Ideally, it would guide users instantly identify your brand, function, theme, and message of the logo in one go. The most advantageous feature, comparing to Latin letters, is the fact that informative message can be conveyed through a single ideograph, while alphabet logos require users to establish an association of brand image in advanced (like f and t). We can see the power of implying in ideographs is a really great example of ‘pre-attentive’ function of ideograph, in fact, there is an overwhelming trend of this approach among Chinese tech firms. Perhaps they have ecologically proven the claim through the process of ‘customer selection’ in a free market?
‘Pay(ment)’, ‘Bean’, ‘Know(ledge)’, ‘Seat’, ‘(Treasure) Hunt’, and ‘Vocabulary’
We have touched on the subject of how precise and short in length this language can be, so it can be expected to see some tremendous space savings when the interface is presented in Chinese. Back to the days of Pokémon on Game Boy when bitmap was literally bit-map, I can still recall how uncomfortable it was when switching from Chinese to the English version and having to make sense out of SP.DF, PAR, ᴾkᴰx, and so on. In the recent App project of our gaming community, I got my hands on mobile-end design for the first time. The déjà vu hit me again as I realised how compressed everything is in Chinese. Taking the iconic Navigation Drawer for a comparison, the content distribution in Chinese is so packed that it takes only half the space and leaving a blank field of nothingness. I had to consider narrowing the width of this component in the guideline if we want our content to be evenly distributed. This is not an extreme case and there are awful lots of instances where official Material Design guidelines have to be modified when targeting Chinese audience.
Thanks but no thanks, I will go with a more compact drawer
It’s not a news that many Chinese apps have created their culturally-specific elements and components, Mr Grover did a far better job than me summarising these interesting cases. However, people tend to attribute all differences to the Galápagos syndrome, could it be that some of these are not a result of isolation but rather caused by features of the language. I do find one of the widely-used components – Bottom Navigation Menus in the chat particularly unseen of in Western apps, this is a component that replaces the input field in the chatting scenario with three uni-width menus. Now you don’t need to know many background theories, but the Chinese language is otherworldly efficient in encoding a large amount of information into a set phrases of four characters. Which you might have realised when looking at the Chinese equivalent of ‘Deals and flash sales’ & ‘Top threads’ entries in the drawer menu above. Anyway, this fact leads to the component where most of the function descriptions can be fitted into a uni-width menu button.
Bottom Navigation Menus in the chat – Image Credit: Dan Grover
Pride comes before a fall, we know that English has 26 alphabets and it takes less than an afternoon to draw a nice set of them, but when it comes to Chinese, the number goes to more than 20,000 or even 50,000 if uncommon ones are included. Unfortunately for the font designers, when a meeting was held to decide how should we represent Chinese in computer output, people went with the solution that requires ‘50,000 or so characters have to be designed individually as a glyph just like from a to z’ (LIGA, ligature) instead of ‘characters can be formed from free combinations of radicals’ (CCMP, Combining Character).
The first four rows of Latin and Chinese Unicode block, the latter is not even capable of counting from 1 to 10
As a result, Chinese, Japanese, and Korean fonts (CJK fonts) are shockingly expensive to engineer. Font projects require so many resources that even companies like Microsoft and Apple consider them as a burden. It is until the April of this year (2017) that we finally meet the very first Chinese serif font ‘Noto Sans Serif’ with multiple font weights, which is at least hundred years later than Latin. Generally speaking, designers face very limited font selection and even fewer if the quality expectation is high. Sadly, there will not be any qualitative improvement to the situation in the foreseeable future and this burden is going to remain for every party behind the screen.
We created a designated font for our last project ‘keylol’ and the cost was indeed unaffordable to small and medium businesses
Should I do something about my workflow then?
At this point, it is no doubt that there are fundamental differences between Chinese and English, which have, to a certain extent, installed unique user behaviours in east Asian countries. But it seems rather unworthy to invest in an exclusive design guideline just to optimise the user experience of this specific group of people, unless they are the sole audience of your product. For now, simply following the mainstream design guidelines will easily score an above average user experience. If you are willing to pay the price, there will always be room for improvement and it will be absolutely welcomed, just don’t expect too much of cost efficiency by doing so.
Behr, F., Fossum, V., Mitzenmacher, M., & Xiao, D. (2003, March). Estimating and comparing entropies across written natural languages using PPM compression. In Data Compression Conference, 2003. Proceedings. DCC 2003 (p. 416). IEEE.
Jianqiao Ge, Gang Peng, Bingjiang Lyu, Yi Wang, Yan Zhuo, Zhendong Niu, Jia-Hong Gao. (2015). Cross-language differences in the brain network subserving intelligible speech. Proceedings of the National Academy of Sciences, 112(10), 2972-2977. doi:10.1073/pnas.1416000112
Siok, W. T., Perfetti, C. A., Jin, Z., & Tan, L. H. (2004). Biological abnormality of impaired reading is constrained by culture. Nature, 431(7004), 71-76. doi:10.1038/nature02865
Shannon, C. E. (2001). A mathematical theory of communication. ACM SIGMOBILE Mobile Computing and Communications Review, 5(1), 3-55.