Re: Emoji characters for food allergens

From: Asmus Freytag (t) <asmus-inc_at_ix.netcom.com>
Date: Mon, 3 Aug 2015 12:38:06 -0700

When it comes to symbolic and / or pictorial representations, there are roughly three kinds. I'm going to use "symbol" here for both symbols and emoji when the latter are used in contexts where there also are (or could be) symbols - and I don't limit symbol to the subset of "encoded" symbols.

• Regulatory or otherwise standardized symbols
• Conventional symbols
• Informal symbols

When it comes to encoding symbols of the first group, the status of the symbol (or picture) is defined externally. There's a standard for laundry symbols, Unicode could decide to encode them. I'll make no judgment here whether that would be desirable. The point I want to focus on is that Unicode should never contemplate to invent extensions to that set, or its own set of laundry symbols. It would be equally inappropriate to allow third parties to "reify" their set of laundry symbols by getting them encoded while the official set is not.

Conventional symbols include things like the skull and cross bones. The image is used in many contexts and has a variety of meanings, which are defined conventionally. These conventions don't have to be very tight: an emoji version of the symbol might equally likely refer to carnival pirates as to danger.

Many emojis and symbols are more informal than that. They either stand in for the object depicted, or for a concept related to the object. A beer mug may refer to "one beer", or to "let's have a beer", etc. In user interfaces a clock face could refer to time, to the concept of waiting, or the concept of "past" (as for items in a search history). Such uses border on conventional, so there is a gray line.

There's a separate classification that's somewhat independent, and divides symbols into by how recognizable they are. If want wanted to divide them into three sets one would get

• iconic
• specific
• somewhat hard to identify

Some shapes and images are highly iconic. They are easy to recognize no matter whether rendered symbolic or realistic. And most users are familiar with most of them.

Some are merely recognizable. Perhaps an image of an object with a distinct shape. Many users are familiar with the symbol or the depicted object, and the typical rendition on a screen is also recognizable.

The there are symbolic or realistic renderings that are less distinct, more easily confused with similar shapes or images, or simply not familiar to most users. (Even highly standardized symbols, like many math symbols, are unfamiliar to most users unless widely read in mathematics -- so this has nothing to do with the degree of standardization, or degree of abstraction in the depiction).

For something like allergens, if there's a need to represent them symbolically/pictorially, it would foremost be as regulatory images. Images with precise limit on their depiction, highly iconic designs that make it hard to misidentify them. And, by virtue of being regulated, they could be forced to be present widely so people would encounter them constantly and become and remain familiar with them.

The least useful thing to be done would  be to encode poorly chosen pictorial rendition of various food ingredients in the hope that users might get the idea that they stand for allergens. Items on the proposed list were not highly recognizable - some I would class in the "hard to identify" group. And, being informal pictures (emoji) they would have no commonality in appearance, negating the benefit of being able to scan for a simple, fixed shape when trying to assess a food item as safe to eat.

Unicode would be very wise to avoid encoding anything for the purpose of depicting "allergens" -- however, as pictorial renditions of food items are popular for other reason, I think some of the proposed emoji could qualify for "generic" use as "food item x". If someone wants to (mis-)use these for allergen information, that should be something that the standard remains firmly ignorant of and something that the Consortium doesn't encourage or endorse.

A./

Received on Mon Aug 03 2015 - 14:39:07 CDT