In defence of the dollop


Google is killing the blob – the yellow character that has personified Google’s version of emoji for years. It will soon be replaced by new designs that are more consistent with other companies’ emoji.

I had seen rumours before that the blob might be going away at some point in the future. And I was disappointed when I heard those rumours. But when I heard the announcement that it was definitely being replaced, I was surprised at just how sad I felt about it. And I’m not the only one.

Why am I so attached to this blob? Why are some people so sad to see the old emoji designs go away, and so disappointed with the new designs? What does this say about how we use and understand emoji? And is there anything to learn about letting third parties own the symbols we use to communicate with each other?

Design as personal memento

Some people think the blob is adorable. Others think it is ugly.

For me, the blob is a character that my wife and I have joked about since we first met and happened to use Google Hangouts to chat. We called it a “dollop”. We both thought it was cute. The dollop, and its “dollopy” attitude to life (as we imagined it) entered our personal lexicon – that private, shared vocabulary of references and in-jokes that every couple or friend group develops.

Years later, we still point out our “dollopy” behaviour to each other on a daily basis. It’s partly a joke, but it’s also a shorthand that lets us say “I understand how you’re feeling”. If my wife has a busy day at work, she might tell me that she’s feeling “a bit stress-dollop”, and I understand this means a specific feeling that’s different from “I’m feeling a bit stressed”.

Similarly, we’ve given each of the dollop emoji our own, more detailed, more specific meanings. When we use the “crying” dollop in a chat, it doesn’t just mean a vague “very sad” – to us, it means a specific kind of disappointment.

Obviously, our attachment to this cartoon character is on the extreme end of the spectrum. But everyone does their own version of this, to a greater or lesser extent. Every practical act of communication imbues every word, every symbol – and yes, every emoji – with personal, contextual meanings. We individualise them, and give them emotional meaning beyond their raw form.

To pretend otherwise is to limit the potential of the tokens we communicate with. And it understates the risk of giving companies like Google control over them.

Ambiguity is unavoidable, and that’s a good thing

One of the main goals of Google’s redesign was “to avoid confusion or miscommunication across platforms”. They cite a study that shows people can have wildly differing interpretations of the meaning of different designs of the ‘same’ emoji. The “grinning” emoji as designed by Apple is often interpreted as expressing a negative emotion, while Twitter’s version of the same “grinning” emoji is often interpreted as expressing a positive emotion.

They ignore, however, another finding of that study: that people can have interpretations of the same design that differ almost as much as their interpretations of different designs:

“We found that if you send an emoji across platform boundaries… the sender and the receiver will differ by about 2.04 points on average… However, even within platforms, the average difference is 1.88 points.”

Even if every company standardised on one design for all emoji, there would still be misunderstandings. Is this a failure of the current designs? Could we fix this misunderstanding with better design? Or, to put it more strongly: could we ever design an emoji that reliably evokes the same interpretation in everyone who views it?

No. We must accept that illustrations and symbols will unavoidably be interpreted differently by different people in different contexts. But that ambiguity is not a bad thing.

Words are flexible, slippery and prone to misinterpretation. Yet somehow we use them to successfully communicate. Sure, sometimes it takes extra effort to clarify a complicated point. And it is usually easier to communicate with someone with whom you have conversed before, whose style of speaking or writing you are familiar with.

Some of the hardest people to communicate with are the ones who are convinced that what they say is clear and unambiguous. Understanding that you are not a perfect communicator is the first step to communicating better. Ambiguity is a jumping-off point for a useful negotiation about meaning – whether that negotiation is spoken or unspoken.

And so it is with emoji.

Friends, partners, and teams create their own consensus about the meaning of ambiguous emoji. This is not a failure of emoji. Rather it is a strength that helps them carry more meaning.

When a new work colleague sends me a message with a smiley at the end, I know it means “this message has some more nuance than is evident from the text alone”. I might even guess that the nuance is that they are slightly joking, or that they are happy about something. But I don’t know precisely what they mean, because I haven’t yet learned how they use that sign.

By comparison, when a particular friend of mine sends me the same smiley, I know that they mean “I’m asking for your help with something, which makes me feel slightly awkward”. Our shared history of conversation, and our knowledge of each other, has imbued that symbol with a very specific meaning.

The false god of consistency

As well as striving to eliminate inconsistency between their emoji and other companies’ emoji, the team at Google set out to reduce internal inconsistency within their own collection of emoji designs.

One of the examples they give is that the “koala” emoji previously showed a whole koala, whereas now it shows you just the koala’s face. Since many of the other animal emoji also just show the animal’s face, the “koala” emoji is now more “consistent”.

Why is this a good thing? According to Google, inconsistencies in their emoji set made it “difficult to quickly scan the keyboard to find the right emoji”.

But which of these koala illustrations is actually easier to pick out quickly from a line-up? The one showing a koala clinging to a tree in a typical koala pose? Or the face-on illustration whose outline is basically the same as the outline of every other animal face?

The koala is a small example of what makes Google’s new design direction unsatisfying. The underlying problem with the new design is that it makes consistency the goal.

Why is this unsatisfying? Because consistency is a means to an end. It should never be the goal itself.

Suppose you are creating two financial calculators. It might make sense to design them so that their appearance and behaviour are consistent. That way, a user can take what they’ve learned about how to use one of the calculators, and apply it to get to grips with the other calculator more quickly. The goal is to reduce the burden of learning two different tools; consistency is a design method we can use to achieve that goal.

But in the case of the koala, we see that consistency makes items harder to distinguish and identify, not easier. If the goal was really to make emoji easier to distinguish, then consistency was the wrong tool.

Good design embraces human messiness

The set of emoji we have today is a glorious mishmash of accidents of history, clashes of culture, practicality, and fashion. There’s no reason that we should expect to find a way to design a ‘consistent’ look that adequately expresses all of them.

Google’s old emoji had non-human dollops to depict most of the abstract emotions, and little humans to depict things like jobs. That’s inconsistent. Their new emoji use human faces to depict both of these things. That’s more consistent.

But what goal does this consistency achieve? Why should the approach for communicating an abstract emotion like ‘confused’ have to be consistent with the approach for communicating a concrete societal concept like ‘the profession of a scientist’? By insisting on consistency for consistency’s sake, the gamut of available design solutions for each emoji is limited, and the potential expressiveness of each design is reduced.

Emoji are not icons

Another technique the Google team used to achieve visual consistency was to base their designs on a ‘grid’ system, similar to the grid system used to design app icons in Material Design.

But emoji are expressive illustrations, not functional icons.

Icons represent, or stand in, for a specific item. The email icon in my dock does not represent the concept of email in general, but rather a specific email client application. The Photoshop icon, for all intents and purposes, is Photoshop. It doesn’t really matter much what the Photoshop icon actually looks like, once I’ve learned that it represents Photoshop.

Emoji are different. The ‘sad’ emoji does not represent a specific sadness, nor a specific instance of being sad. The sad emoji is the sad emoji. How it looks is what it is. The users of the emoji get to decide precisely what it represents for them. And the different design interpretations of ‘sad emoji’ just add to the richness of what it can communicate.

Why is it so tempting to approach the design of emoji as if they were icons? Because the Unicode standard for emoji defines them in terms of a descriptive piece of text. In the same way that the Photoshop icon represents Photoshop itself, it is tempting to think that your design for the emoji defined as “STAR-STRUCK” is an icon that represents the text “STAR-STRUCK”.

But it is not. The user never sees the text “STAR-STRUCK”.

Typically, a user selects an emoji by visually scanning through a list of the emoji themselves. The closest they might get to a textual representation is the text-based emoji search feature that some software has. But even there, the user’s model is usually that they are searching for a visual element inside the emoji they want (e.g. “cat” or “water”), not that they are finding an emoji that directly represents the text they have entered.

This seems like a subtle distinction to make. But it’s important because it gives us insight into the user’s understanding of what ‘digital material’ emoji are made from.

Learning the physics of digital material

As children, we learn about the physics of materials in the real world. We learn that a piece of wood can stand up by itself, in a way that a quantity of water cannot. We learn that some materials, like dough, can have their shape easily changed. We learn that ink on paper is relatively permanent, whereas chalk on the pavement changes when the rain comes.

When people start communicating in the digital realm, they learn that there are different materials here too, and that they behave in different ways. When I send a picture to someone, it usually arrives unchanged, verbatim. But when I send a text message, I quickly learn that the recipient might see the lines wrapped in different places, and in a different typeface from the one I see.

More inconsistencies. Perhaps it would be better if all our digital materials could arrive unchanged, looking exactly as we sent them. Indeed, some propose this as the ‘solution’ to emoji inconsistencies between different platforms – every platform should just agree to use the same design for every emoji.

But that’s not how this digital material actually works. It incorrectly characterises the behaviour of emoji. Emoji are not little pictures that will be transmitted verbatim. They are more like text. And one of the ways that people learn this true nature of how emoji work is by experiencing inconsistent designs for the ‘same’ emoji.

We should have more faith in the intelligence of our users. They can cope with a richness and variety of emoji designs.

Yes, that richness can be confusing. And yes, it sometimes requires negotiation about what we meant to say. But that very richness and variety are what make emoji versatile, funny, emotional, and loved.

Practical tools versus cultural tokens

When my bank changes the visual design or organisation of their online banking tools, I am inconvenienced for a while. I need to re-learn how to use the tools, and I may even find that the old version was more convenient for the tasks I tend to use it for.

Ultimately, though, a change in the design of a tool is rarely more than an inconvenience.

But a change in the design of a cultural token or sign causes a change in our vocabulary and communication.

These days, much of the software we use and the data we create are held by third parties. Because of this, these changes can be forced on us for future communications. And they can retroactively change personal artefacts from our past.

Suppose a friend, relative or partner had sent you handwritten letters that were important enough for you to keep as a memento. Imagine that many years later you go back to read them again, and find that every instance of the word “love” has been changed into ❤️.

You haven’t exactly lost anything, but how would that make you feel?

Tech companies generally do a bad job of communicating change to their users. Of course, they fervently believe that their new version is better than the old, and this affects how they talk about it. But, in the case of tools, they could have more empathy with the inconvenience that change brings to users. And in the case of cultural tokens, they must acknowledge the effect their changes have on the personal communications and relationships of their users.

I would have been comforted to read that some Googlers were, like me, a bit sad to see the dollops go.

Sensitive to change

From all of the above, you might think I’m arguing that emoji designs should never change. But I’m not against change in principle. And Google’s emoji team has been at the forefront of some important changes for the better.

I am, though, disappointed that Google has thrown away a characterful design that has breathing space to be expressive, that embraces human messiness by its internal inconsistency, and that teaches us about the true nature and richness of the digital material of emoji through its courage to be different from other companies’ designs.

And I am sad that I won’t be able to send dollop emoji easily anymore, and that all my old conversations will be altered by this change. But that’s really more to do with allowing Google to host all my conversations, and thereby have complete control over how they appear. I’ll think twice about letting that happen again.

Farewell, dollop. I’ll miss you.