Iconography

A book without words, recounting a day in the life of an office worker, told completely in the symbols, icons, and logos of modern life.

Some of my best friends are designers. But I think we can all agree that - however well-meaning - they can be a little obsessive. Whether it is fretting over tiny details, or trying to align to a grid which doesn't exist, or spending time removing useful affordances in the name of æsthetics - they always find a way to make something prettier at the expense of usability.

Google used to have some beautiful logos for its apps. Each had a distinct shape, style, and colour. Then, someone decided that they all needed a consistent visual language. And this mess was born.

↬ welcome to the best part of the unicode spec ↫

Months ago, I was pondering ways to improve my photoessays. My posts had focused almost purely on the photos and the stories. I felt that describing the gear used would be an improvement since gear is a small but integral factor.

I was considering listing out cameras and lenses at the end of each post. That’s when I remembered my previous drawings of cameras. I decided an illustration with a legend would be a welcome addition.

In March 2021, I went through a fun self-imposed experiment: no emoji for 2 weeks. Not on social media, not in private messages, not even as Slack or Discord reactions. No emoticon either: the goal was to communicate without illustrations, only with words. I did a semi-rigorous (a.k.a. half-assed) diary study, taking notes on my feelings and behaviour.

If I were tasked with finding an icon that can be used to label an AI-related feature in a UI component, I’d likely pursue a set of icons more closely tied to the feature itself, such as the sparkle-donning envelope for AI-flavored email features. The sparkle is what suggests automation and whatever it is applied to it is the feature. Sparkles alone can mean too many things (and reek of marketing cruft).

...Some of this isn’t specific to AI but to iconography in general. The detail needs to be clear enough to convey meaning, but not so detailed that the meaning is lost when downscaled. The elements need to be recognizable at first glance but also avoid crossing into well-known commodities, like how the helix shape closely resembles the logos of React and The Big Bang Theory. The subject matter has to steer clear of creepy connotations or else suffer the fate of Uncanny Valley. These are all considerations to take into account when working with icons of any kind.

Apple’s announcement of “dark mode” icons has me thinking about how I would approach adapting “light mode” icons for dark mode.

...The white-background icons simply become black-background icons. Maps utilizes a dark mode color palette from the app itself, Weather turns the sky black, but oddly keeps the sun rather than switching to the moon. This could be a rule Apple enforces only for themselves, where their app icons won’t change shape, only coloration. The Photos petals are now additive color rather than subtractive.

Unfortunately, some icons appear to have lost or gained weight in dark mode. For example, the Settings gear didn’t change size in dark mode, but it appears to occupy less space because the dark circle around it blends with its background. That makes it appear smaller than the Find My icon, which now looks enormous next to FaceTime. This is a remnant of some questionable design choices in iOS 7 that have lingered now for the last decade.

Everett’s core idea is that language should properly be thought of as an invention rather than an innate human capability. You might ask: who invented it? Who shaped it? Lots of people, collaboratively, over a long time. In a word, culture. As Everett notes in the preface, “Language is the handmaiden of culture.”

...Everett, in the tradition of Peirce, distinguishes between various different types of signs. The distinction is based on (i) whether the pairing is intentional, and (ii) whether the form of the sign is arbitrary. Indexes are non-intentional, non-arbitrary pairings of form and meaning (think: dog paw print). Icons are intentional, non-arbitrary pairings of form and meaning (think: a drawing of a dog paw print). Symbols are intentional, arbitrary pairings (think: the word “d o g” refers to a particular kind of real animal, but does not resemble anything about it).

Everett argues that symbols did not appear out of nowhere, but rather arose from a natural series of abstractions of concepts relevant to early humans. The so-called ‘semiotic progression’ that ultimately leads to symbols looks something like this:

indexes (dog paw print) -> icons (drawing of dog paw print) -> symbols (“d o g”)

...Why did we end up with certain symbols and not others? Well, there’s no good a priori reason to prefer “dog” over “perro” or “adsnofnowefn”, so Everett attributes the selection mostly to cultural forces. Everett suggests these forces shape language in addition to practical considerations, like the fact that, all else being equal, we prefer words that are not hundreds of characters long, because they would be too annoying to write or speak.

This is a clever attempt to write a spring couplet (chūnlián 春聯), not in the usual Sinoglyphs / Chinese characters, but in pictographs.

...Of all the symbols, the lamest is the one for "nothing", which is a real cop-out, simply using the character wú 無, which means "nothing".

So now I must ask, why is it so hard to depict "nothing" (they could perhaps have drawn an assemblage of five objects for the near homophone wǔ 五)?

The abstract idea of "nothing" is hard to depict in a morphosyllabic script like Chinese, so the devisers of the oracle bone script borrowed the glyph for "dance" to stand for the homophonous "there is not; nothing"). Later, to distinguish the latter meaning from "dance", they invented another character by adding a component at the bottom, thus wǔ 舞.

The logo consists of a rhombic geometry which was inspired by Morioka’s drawing which he brought to the first meeting with Takram. During the process of design development, the team explored many different shapes and motifs other than the rhombic, but in the end came back to the very origin. In fact, the rhombic shape embraces two meanings, “an open single book” and “a single small room”. The first message was the vision team shared through the design process, and the second message was later proposed by Takram in order to make the geometry more connected to the character of the space, which Morioka often emphasised unconsciously.

An image can take the place of a word in a proposition.

The thing that makes Google’s later icons extraordinarily incompetent is that the old icons were consistent in terms of style, colours, shading, etc. The new icons are worse because they are less identifiable.

Google and Apple have been driving towards greater uniformity and less variation in their software design languages over the past few years, but their uniformity doesn’t mean they’re genuinely more consistent designs

For example, the older icons tended to have a consistent approach to using metaphor, whereas the newer icons are basically deconstructed abstract shapes that make no sense UNLESS you actually remember the older icons they’re referencing

These are just bad designs. Plain and simple.

The section sign (§) is a typographical character for referencing individually numbered sections of a document; it is frequently used when citing sections of a legal code. It is also known as the section symbol, section mark, double-s, or silcrow.

In 1923 Kandinsky proposed a universal correspondence between the three elementary shapes and the three primary colors: the dynamic triangle is inherently yellow, the static square is intrinsically red, and the serene circle is naturally blue.

The series ▲■● represents Kandinsky’s attempt to prove a universal correlation between color and geometry; it has become one of the most famous icons of the Bauhaus. Kandinsky conceived of these colors and shapes as a series of oppositions: yellow and blue represent the extremes of hot/cold, light/dark, and active/passive, while red is the intermediary between them. The triangle, square, and circle are graphic equivalents of the same polarities.

The main problem, I think, is that no one knows what "Verified" means.

If I were in charge (which I'm not) there would be various types of ticks.

🤖 is a bot
🆔 proved their legal identity
🏭 is run by a brand
⚖ is run by a government department
👮 Official law enforcement
😎 Celebrity

And so on.

An icon is a symbol equally incomprehensible in all human languages. There's a reason why humans invented phonetic languages.