Historical Perspective – Implicit Cultural Knowledge
There are many “standard” AAC symbol sets, from Picture Communication Symbols (PCS), to Minspeak/Unity, to Widget Symbols, and more. What standard symbols sets have going for them is that they’ve been around for a long time – and not much more. At the ATIA 2020 conference, we stopped by a vendor’s booth where we saw an AAC device with, what appeared to be, a new graphic symbol set. We asked the sales person what research they called upon for their symbol set. He simply said, “we have a person in development with graphic design skills”.
Boardmaker (a leading proponent of Picture Communication Symbols) provides a history of the symbols:
“One of the earliest companies in the Assistive Technology field was Mayer-Johnson, which started as a mail order company focusing on augmentative communication products. Roxie Johnson, co-owner of the company with her husband Terry, had been working as a speech therapist with special needs children in the public-school system and with adults in a hospital setting. She saw the need for a set of picture symbols that limited-speaking or non-speaking persons could use for communicating. She developed the first book entitled The Picture Communication Symbols (PCS) that included 700 picture symbols. The symbols were designed with thick lines, so they could be photocopied to support multiple users.”
These two stories have two key elements in common. Both have a single or small number of individuals rapidly generating images to represent a small number of words. Both involve designers, with rich cultural knowledge, unconsciously and unavoidably embedding that cultural knowledge in the images they create. The result is a hodge-podge of graphic images with no underlying rules driving them, and so, no predictable relationships between related words. The image below still exists in some implementations of the PCS set and demonstrates the amount of unconscious cultural knowledge that can be buried in a graphic:
In order to connect this image to the concept “to want”, you must first recognize that the smaller image is a stereotypical “criminal” – with a cartoonish mask and tam hat. The small image then becomes a “wanted poster”. Why the larger individual in the image is reaching for the wanted poster and why he is wearing a black and white horizontally striped shirt is unknown – though it does invoke images of prison garb. Roxie Johnson made the inexplicable leap from the concept “to want”, to the concepts “wanted criminal” and “wanted poster”, when she came up with this graphic. Would you have made these connections without this description? Can you imagine a child with very little cultural knowledge making these connections?
Choosing a graphic symbol set for your AAC device simply because it comes with “lots of pictures”, and no evidence that the pictures map easily to their referent, is misguided.
Some images (especially those that represent less tangible referents) are often “drawn” from a wholly separate language – sign language. The images below are intended to represent the concepts “to do”, “finished” and “more”:
 |
 |
 |
If an individual is being taught to use sign language at the same time as they are learning to use their AAC device, there may be value in taking this approach – but then, why not use images of signs throughout? Why only when illustrating a challenging concept? There is no inherent relationship between these images and their referent concepts – except by first leaping to an entirely different language.
By contrast, Blissymbols have an underlying architecture to ensure that related concepts contain similar elements and structure. The symbol set is generative but Blissymbolics has a standards body to ensure that an official symbol set is curated. With Blissymbols, you get a symbol set that has been thought-through, and tested, for consistency and inherent meaning across multiple cultures. Best of all you get a high-quality symbol set for free!
Minspeak and Semantic Compaction
According to Wikipedia, “Semantic compaction, (Minspeak), conceptually described as polysemic (multi-meaning) iconic encoding, is one of the three ways to represent language in Augmentative and alternative communication (AAC). It is a system utilized in AAC devices in which sequences of icons (pictorial symbols) are combined in order to form a word or a phrase. The goal is to increase independent communication in individuals who cannot use speech. Minspeak is the only patented system for Semantic Compaction and is based on multi-meaning icons that code vocabulary in short sequences determined by rule-driven patterns.”
Semantic Compaction was created by Bruce Baker as a way to provide AAC users a more efficient method for constructing phrases and sentences with their AAC device. The Minspeak website describes the process as: “Minspeak icons are combined in short sequences – usually 2 or 3 icons – in order to say different words. For example, you say furniture by sequencing BED with the NOUN key, sleep with the VERB key, and tired with the ADJECTIVE key.”
As a pictographic language, Minspeak suffers from all the issues described above.
Why would an old lady looking into a medicine cabinet (or is that a kitchen cabinet?) represent the concept of “noun”? How is a child to recognize the fellow in the green uniform as a “generic worker” and why is he carrying a bucket and a hammer (if that’s a hammer)? Then the child must make the leap from this symbol to “verb”. No doubt the paint brush is being used as a metaphor for altering or modifying the look of something and, therefore, represents an adjective. But why do this with a “black-and-white” icon?
As a meta concept, semantic compaction has to balance on this unstable, pictographic base.
The second-order problem comes from trying to build unique concepts from a small set of icons. Apparently the “bed” icon, when combined with the “noun” icon, represents the concept of “furniture”. What does the “chair” icon, in combination with the “noun” icon, represent?
Is it possible to do anything else in bed besides sleep? We leave that question to the reader…
Blissymbols utilize a vaguely similar methodology – that of taking a base concept and generating new concepts by superimposing or combining it with other concepts. More directly comparable, Blissymbols create many verbs by adding an activity indicator to a base symbol. However, the Blissymbolics Communication International standards body strives to ensure that there is minimal ambiguity when this is done.
Stewardship and Cost
As discussed above, it’s important that a graphics set be based on rules so that the graphics build upon one another. It’s also important that each graphic be as simple as possible. It should include only enough content to make a clear and simple connection to it’s referent without requiring any assumed, or unintended, cultural knowledge. It’s difficult, maybe even impossible, to accomplish this task without a dedicated body of individuals who are charged with ensuring that these goals are met. It’s even harder if one starts with a significant base of graphics for which these goals were not a priority at the time.
What do the “standard” AAC graphics cost? It varies significantly:
| Symbol Set | Cost | # of Symbols |
| Picture Communication Symbols | $400 | 45,000 |
| SymbolStix | $130/yr | 60,000 |
| LessonPix | $36/yr | 40,000 |
| Widgit (Rebus) Symbols | $?? | 17,000 |
| Minspeak/Unity** | n/a | n/a |
** Minspeak is available only by purchasing an AAC device from PRC
Blissymbols are carefully curated, free to use, and include over 6000 “blessed” symbols New concepts are added to the official set each year. Because Blissymbolics is a generative language, new concepts are constantly being created by the users themselves.
Other Aspects
In a world where “research-based” practices are preferred or possibly even required, let’s look at how much research (in the form of peer-reviewed journal articles) is associated with each of the “standard” symbol sets that we’ve identified. The Project Core tactile symbols are included here because they’ve originated from a university and therefore carry a certain credibility by that association.
A quick academic journal search found the following counts of articles since 2001. PCS have been around for many decades and are the most used symbols in AAC devices so it’s natural that they have the most associated research regarding their effectiveness. Widgit Symbols have a similar longevity (mostly as Rebus Symbols). The newer symbol sets, like SymbolStix and LessonPix, have no “academic research” evaluating – pro or con – their effectiveness.
| Symbol Set | Peer-Reviewed Journal Articles |
| Picture Communication Symbols | 361 |
| SymbolStix | 0 |
| LessonPix | 0 |
| Widgit (Rebus) Symbols | 65 |
| Minspeak/Unity | 6 |
| Project Core Tactile Symbols | 0 |
| Blissymbols | 211 |
It’s interesting that there are no peer-reviewed journal articles supporting the use of the Project Core tactile symbols, though they are accepted for use by many SLPs and TVIs because of their association with the University of North Carolina.
Conversion to Tactile
Finally, everything we’ve said up to this point is moot if a graphic representation can’t survive the conversion from a visual modality to a tactile one. There’s been a lot of research on this topic. Not necessarily to address the needs of students trying to learn language but, more commonly, to help individuals who are blind or visually impaired navigate a world where so much information is encoded graphically.
The challenge of converting graphic images to tactile equivalents is captured perfectly by this statement by Philippe Claudet in his article in The Journal of Blindness Innovation and Research titled “Designing Tactile Illustrated Books”:
“most tactile pictures are visual ones made in relief by sighted people for whom it is very difficult to forget their sight”
So, much like graphic images that unwittingly embed critical cultural knowledge, sighted designers often, unwittingly and unavoidably, embed visual cues in their tactile symbols. There are often visual cues in graphic images that are critical to their understanding and are difficult to translate to a tactile-processing mode. One of the most popular cues is the use of color to indicate that two separate parts of the image are the same object or made of the same material. There is no easy way to convey this same information tactilely without fundamentally altering the image.
Jiangtao Gong, et. al., make these recommendations in their article “I can’t name it, but I can perceive it” Conceptual and Operational Design of “Tactile Accuracy” Assisting Tactile Image Cognition in The 22nd International ACM SIGACCESS Conference on Computers and Accessibility:
(1) Avoid using obvious 3D perspective effects such as “foreshortening” or showing multiple faces of an object at the same time. Try to use a parallel perspective or cut slices of the object rather than use a 3D perspective.
(2) Avoid breaking the main contours or occluding the important details of the objects in the image. The overall outline of the object should be kept intact, and the important details should be fully displayed in the blank area to avoid confusion between the foreground and the background.
(3) Avoid using undersized or oversized tactile images, which are too small or too large relative to the physical scale and can impair the tactile recognition.
(4) Try to preserve the symmetry information while depicting a symmetrical object in the tactile image.
Here again are the images that Minspeak uses to represent the concept: “furniture”:
The image of the bed uses color to represent three different layers of cloth and shading to represent a change in dimension for the red blanket. The color brown is used to imply that the bed frame is constructed of wood. The foot rail of the bed is longer than the head rail because visual perspective demands it, though in reality, the rails are the same length. In the picture on the right, the woman’s head occludes the cabinet door so we see two lines below her chin that, as sighted individuals, we know are part of the door but would be mysterious if encountered only tactilely. The woman is as almost as large as the entire bed, but we know that her head would only be as large as the dent in the pillow. (Speaking of which, what makes the dent in the pillow a dent? A subtle line and an even more subtle bit of shading.)
It’s not our intent to malign the Minspeak icons. They weren’t developed with the intent of becoming raised tactile symbols. We’re just saying that using them, or some other “standard” set of complex visual graphics as the basis for a tactile graphic set, is fraught.
Just out of curiosity, what is the Blissymbol for the concept “bed”?
The image is black and white, without perspective, emphasizing the symmetry of the object along with the one feature that distinguishes the head end of the bed from the foot end – a simple representation of a pillow.
While Blissymbols were not developed for use as a tactile language, they have many characteristics that make them the perfect choice. Mick D. Isaacson & Lyle L. Lloyd saw the same potential in their article “The potential for developing a tactile communication system based on Blissymbolics” in Developmental Neurorehabilitation, 2015:
“Due to their simplicity, edges, outlines, and kinesthetic feedback are optimized in raised-line renderings of Blissymbols, which may facilitate recognition of tactile Blissymbols. In short, the conformational characteristics of Blissymbols results in stimuli that have minimal complexity (or low potential cognitive load) and when rendered as raised-lines have characteristics that may increase the efficiency of perceptual/cognitive processes involved in communication with tactile symbols.“