My first design teacher, Philip Burton, used to say, over and over, don't be anecdotal. He would say, when we were presented with a design task in class, most of which were abstract in nature, that we weren't making rebuses. (Paul Rand's famous IBM ad is a design joke and intended to be amusing, not the company's new logo.) What Philip meant about anecdote is that we should not be looking for a literal rendition of what we wanted to represent, but rather a figurative one.
There is a break in childhood development: a point at which you move from concrete to abstract thinking. Some people never get there. Engineers, software or otherwise, use an enormous amount of abstract thinking, but some also often appear rooted in the concrete for everything but math and science. For them, visual and literary aesthetics involve a literal reading: if you say one thing but mean the other, you're not making sense to them. The idea of gray areas, shading, or things that they can't immediately grasp fully are rejected. When I read Mayer's explanation, I feel like I'm peering into the mind of an engineer.
I'll leave a more detailed and excellent critique of the design process and actual design to Oliver Reichenstein of iA, who wrote "Logo, Bullshit & Co., Inc." Yes to everything he said. But I want to make a specific point about type design.
Mayer and her team adapted Optima, one of the first humanist sans serifs: the swell of the letter implies a serif and lacks the more straight and geometric (not mathematical) edges of many grotesk sans serifs, like Helvetica. Designed by Hermann Zapf, one of the greatest type designers of the last century, Optima has a slight softness to it that makes it perfect for non-fiction prose, where a grotesk would seem too severe. Optima, like his Palatino and Melior, are all around us. (Zapf is 94 and still kicking, and I will be curious if he will have anything to say.)
Now, it's not sacrilege to modify a type design; there are licensing issues, but there's nothing sacrosanct about it. Apple famously had Bitstream make an extra-compressed version of Garamond that became its standard font, and which never looked wonderful to my eyes. But to make changes to a face while ignoring the entire history of the written word and coming up on 600 years of movable type used for Western alphabets is ludicrous.
Those born into the age of advanced electronics and the Internet are always inclined to discard the past and start fresh. A tabula rasa is wonderful, but only if there is literally nothing to learn from history. The worst thing about Amazon.com in its early days (I ran the book-information department for six months in 1996–97) wasn't the hours nor disorganization nor some of the personality problems of early hires given too much authority. It was a great place to work in many ways.
The worst thing was that there was an assumption that there were no lessons to learn from anyone in the book industry or with any knowledge of warehouse logistics or shipping. I argue that Amazon wasted billions of dollars and a substantial fraction of its person-hours until about 2001 by reinventing every single thing from scratch instead of carefully examining what worked and what didn't. (Barnes & Noble is the cautionary tale on the other extreme, launching its site relying only on what it knew and had on hand.)
We wanted there to be a mathematical consistency to the logo, really pulling it together into one coherent mark.
If there is one sentence in the entire blog post that tells the whole story, that is it. This shows that not only does she lack an understanding of design — which is fine, it's not where her strengths lie — but that she also doesn't know it; that designers consulted were unable to disabuse her of this ridiculous notion; and that the final result pleased her, when it is obviously flawed in this regard.
The only type designs that are "mathematically consistent" are used for computer-readable purposes, such as fonts developed to be magnetically scanned off checks. All other faces, including monospaced faces used in text editors or for developing software and ones that simulate monospaced or proportionately spaced typewriter letters, are designed for optical consistency.
Designing for mathematical consistency ignores three related factors: that identical widths and shapes appear differently to the eye in different combinations within a letter or glyph; that identical shapes blend together and are harder to differentiate across words and lines; that letters in a typeface are placed alongside each other, and one must adjust to deal with common juxtapositions.
The human eye is insanely good at spotting things that look off. A bad typeface looks bad to people who aren't trained in design, because it lacks a visual fluidity that good type has. The same is true with a logotype, like Yahoo's.
Type evolved from writing with a stylus or brush which evolved from chiseling, carving, and other earlier methods (still employed). Humanity didn't first invent a form of representing speech of ideas and require the mind and eye to catch up. We produced things that were consonant with our ability to interpret them.
Technology has fooled us into thinking that we invent first and train our minds later. Our minds are prefabricated to read type as it has evolved, because type evolved to interoperate with our visual cortex and our neural structure. Our brains haven't suddenly changed.
Ignoring the development of useful things that developed alongside us is arrogant and misguided.
When I look at the Yahoo logo, I have far less hope for its future.