Give a hoot! Mapping (and caring for) the semantic environment

GIVE A
H T!
Mapping (and caring for)
the Semantic Environment

I recently moved to California, and was surprised to discover that people here don’t just
recycle: they separate waste into garbage, recyclables, and compost. In Panama, we had a
single bin where we dumped everything. So my family and I have gone through the process of
learning a new taxonomy of trash.

This is happening because over the past 50 years or so the developed world has rediscovered
the importance of the impact of our interventions in the physical environment. We now
realize that the quality of our lives depends on the quality of our environment, and that we
should do everything we can to curb those practices which damage this environment.
Photo: http://www.ﬂickr.com/photos/foolswisdom/230164685/

Pollution
As obvious as it sounds now, the realization is fairly recent: it’s only come to the fore after
the Second World War, as we started to pay the price of unchecked industrialization. This is a
photograph from the Great London Smog of ’52, which claimed between 4,000 and 12,000
lives and sickened 100,000 people.
Photo: http://bit.ly/1eF5pRR

Los Angeles, 1948
This is the Los Angeles Civic Center in 1948.
Photo: http://bit.ly/1fZHamK

Cuyahoga river
The Cuyahoga river in Ohio, which was so polluted it literally burst in ﬂames — more than
once! This all happened within many of our parents’s lifetimes Disasters like these, the
publication of critical books like Rachel Carson’s Silent Spring, and the counterculture
movement of the 1960s led to a renewed concern with environmental issues.
Photo: http://bit.ly/1iaO4US

Which brings us to this guy. This is Woodsy Owl. He was created in 1970 by the U.S. Forest
Service to raise awareness of the importance of protecting the environment. His slogan —
“Give a hoot! Don’t pollute” — was meant to get people to interact more conscientiously with
their physical environment.
As you can see from my California experience, Woodsy has been successful in his mission, so
I wanted to bring him along to tell you about caring for a different type of environment we
live in: the semantic environment. However, in 1974 the U.S. Congress passed public law
93-318, also known as the Woodsy Owl Act, which protects Woodsy’s image.

WORDSY IA
So I’ve asked Woodsy’s cousin Wordsy to join us instead. Wordsy will be prompting us from
time to time with points to keep in mind as we learn to appreciate and protect the semantic
environment.

Prix fixe
1. Why are we having this little talk?
2. General Semantics
3. The Semantic Environment
4. Craziness vs. Stupidity
5. … in practice
6. Coda
But first things first. Allow me to give you an overview of what’s on the menu for our time
together today. First, I’ll tell you about why I care about this topic, and why I think you should
care too. Then, I’ll talk a bit about general semantics, the field which prompted these ideas.
Next, I’ll explain what a semantic environment is. With this framework in place, we’ll talk
pathologies, especially the difference between “crazy talk” and “stupid talk”. Then, I’ll show
you a way to use this framework to map semantic environments so you can avoid making
stupid or crazy changes. We’ll end with a coda, my own little “one more thing”.

Q: Why this talk?
A: It’s all semantics
So why are we having this conversation? About 12 years ago I was already a few years into my
career as a digital designer, having transitioned from building architecture. I had already read
Wurman and the polar bear book, and self-identiﬁed as an information architect.

I was looking online for my tribe, and came across this organization called the Asilomar
Institute for Information Architecture. I knew some of the folks behind it from mailing lists
and such, but I was particularly struck by the “About us” page in the Institute’s website. It
included a deﬁnition of IA that resonated deeply both with my background and experience.

“The structural
design of shared
information
environments”
It said that IA is “the structural design of information environments”. As an architect, I loved
the idea of designing structures for information environments: this offered suggestions on
how I could apply my background to this new area of practice. However, I was always clearer
on the “structural design” part of this sentence than on the “information environments” part.
What exactly is an information environment? After all, you could show this sentence to an IT
engineer, and she could claim that this is what they do as well. I’ve spent many years trying
to understand what exactly we mean by this.

For most of my career, the dominant medium that I’ve been working with is the World Wide
Web. However, this picture has become more complicated as new information access
mechanisms, such as smartphones and tablets, have become popular and have allowed us to
interact with information in different contexts.
Photo: http://bit.ly/O3hN4W

The structural
integrity of meaning
across contexts
At last year’s Summit, I told you that I believe that one of the most important beneﬁts that IA
provides is the ability to convey meaning across multiple contexts. This statement raises
many questions, especially in light of this concept of designing for “information
environments”, but foremost in my mind are “what is the relationship between this phrase
‘information environments’ and ‘contexts’, in the sense we’ve been using it?”

The structural
integrity of meaning
across contexts
And also, “how can you convey meaning effectively across different contexts?” This word
“meaning” is particularly tricky: things have different meanings for different people in
different times and situations. It’s slippery stuff, and somehow we’re supposed to produce
structures that ensure that meaning is effectively conveyed.
I know many people are turned off by discussions about semantics, and you may be too.
However, I believe it’s important to think about this stuff, because as Andrew Hinton told us
this morning, language is our main building material. So I’ve been thinking about information
environments, meaning, and how we can enable understanding — the transmission of
meaning — through information environments.

General
Semantics
This led me to a ﬁeld called general semantics and a concept called the semantic
environment, which is very close to what I understand information environments to be. I will
share some of what I’ve learned with you today.

Alfred Korzybski
General semantics was started in the 1920s by a Polish-American philosopher and scientist
called Alfred Korzybski. His lofty goal was to develop a framework that would allow us to
alter human behavior in the direction of greater sanity. Because language inﬂuences our
thinking, and thinking inﬂuences our behavior, language is the core concern of general
semantics.
Photo: LIFE Magazine (1948)

Alfred Korzybski
“The map is not the territory”
“A map is not the territory it
represents, but if correct, it has a
similar structure to the territory,
which accounts for its usefulness”
You may have heard the saying “The map is not the territory”. This is Korzybski’s best known
statement; it is one of the main tenets of general semantics. However, he issued a longer
version which is more convoluted but also more useful for our purposes. You’ll note two
things: there is a concern with structural integrity, and there is a concern for usefulness. This
focus on usefulness is important: GS is concerned with the way that individuals interpret
meaning through the way we use language, rather than with the transmission of meaning in
the abstract.
Photo: LIFE Magazine (1948)

Surgeon Patient Statistician
This all sounds a bit ethereal, so let me give you an example. Imagine a patient has gone to
an appointment with his surgeon where he will be told the results of a biopsy. Sitting in on
the case is a hospital statistician who is keeping records on the results of biopsies in the
hospital.
Diagram: J.S. Dubois

“This is cancer”
The surgeon delivers the news: she says, “This is cancer.” Now, these three people are going
to interpret these same three words very differently. The patient is probably devastated,
scared, and anxious. The surgeon is ﬁnding ways of comforting the patient and making
treatment plans. The statistician checks a box and moves on to the next case.

SEMANTICS
The study of SYMBOLS
The ﬁeld of semantics is concerned with studying the symbols that comprise the statement
“this is cancer” and what they mean.

SEMANTICS
The study of SYMBOLS
GENERAL SEMANTICS
The study of HAPPENINGS-MEANINGS
General semantics, on the other hand, is concerned with how the human beings involved in
the situation react to and derive meaning from these symbols, which we call language. In GS,
language is considered a distinctly human ability, one that sets us apart in the world.

Energy-binder
General semantics postulates that lifeforms can be organized into three basic categories
which have expanded in scope over time. Plants, which were the ﬁrst to appear on the scene,
are energy-binders. Their major contribution was ﬁguring out how to transform nutrients and
sunlight into energy via chemical reactions.

Space-binderEnergy-binder
Animals added the ability to move around, thus establishing a more complex relationship
with their surroundings. Because of this, we can think of them as space-binders.

Time-binderSpace-binderEnergy-binder
Our species opened the third category: we are time-binders. Because we have language, we
can learn from the past. Language is central to what we are, and it is what has enabled us to
build and learn and grow from each other and people who came before us. Now, you could
argue that some animals are capable of simple language. However, as far as we know,
animals are not capable of abstraction.

Abstraction enables you to think and talk about things beyond the here and now, beyond the
concrete. For example, you can’t take a photograph of a “musician”.
Photo: https://www.ﬂickr.com/photos/zoooma/2406667119/

Guitarist
Musician
If you want to illustrate a musician, you have to step down at least one level of abstraction to
show something like a guitarist, which is a more concrete type of musician. However, we can
talk about music and musicians as abstract concepts.

Guitarist
Rock guitarist
Jerry García
Musician
Artist
Human
ABSTRACTION
We can easily go up and down levels of abstraction, and we do so all the time; if you’ve ever
searched for stock photos, you know this ability is incredibly useful. However, we have to be
careful, because at higher levels of abstraction, things are more open to interpretation. Also,
we carry inside us representations of the world made from these abstractions. This is the
map Korzybski was talking about. We have to be careful to not be seduced into believing that
the representation and the thing being represented are the same.

As new ideas and things appear in the world, language evolves. We don’t usually learn new
words by looking them up in the dictionary; instead we see them used in different contexts
and little by little ﬁgure out what they are (and aren’t) referring to. To illustrate, let me share
with you a word that has been around for a while, but which I only learned a few years ago.
Image: Rene Magritte, “La clef des songes”

SKEUOMORPHIC
This word “skeuomorphic” has been used a lot in our circles recently. The first time I saw it
was on a blog post by Adam Greenfield. I had no prior reference to it, so it took me a while to
actually get what it meant. After much reading, I now think it means something like: “an
object or design which provides visual affordances by superfluously copying characteristics of
older designs.” As “skeuomorphic design” became a thing we talk about, we developed the
need for a word to describe its opposite.
Photo: http://www.flickr.com/photos/pinkmoose/2611301708/

FLAT?
The word we chose collectively, “ﬂat”, is not an obscure word like “skeuomorphic” — it is
already widely used and infused with meaning. So it brings with it all sorts of associations
that in my opinion are unhelpful when we’re trying to describe UI designs.

FLAT?
For example, the iOS 7 UI — one of the more famous “flat” designs — is not flat at all. As a
matter of fact, layering in the z-axis is one of its core differentiators. It does it a disservice to
call it flat, but it’s not skeuomorphic either. When describing iOS 7, we are trapped by a
duality in language.

SKEUOMORPHIC FLAT
SKEUOMORPHIC MODERN
SKEUOMORPHIC HONEST
SKEUOMORPHIC PLAIN
SKEUOMORPHIC TASTEFUL
It’s worth noting that there’s nothing about the term “skeumorphic” that suggests that it sits
at the opposite end of a continuum with “flat”. I can think of many other terms that would fit
just as well. However, the fact that “flat” became the accepted term has reinforced the use of
a particular esthetic over other possible ones. In other words, the words we use to describe
and categorize things have an impact on not just how we see the world, but also on how we
shape it.

Many problems in the world are ultimately problems of classiﬁcation and nomenclature.
Society regards as “true” those systems of classiﬁcation which produce the desired results,
even though those results and objectives may be atrocious. What we name things — and
people — has a huge impact on our experiences and interactions with the world and with
each other.
Photo: http://bit.ly/1hdQnDm

WORDSY sez… GIVE A HOOT!
Remember: when you name
something, you constrain
what it CAN and CAN’T BE!

The semantic
environment
So now we’ve gotten through some of the basics of general semantics. Let’s focus on this
idea of the semantic environment, which came out of GS.

Neil
Postman
Crazy Talk,
Stupid Talk
(1976)
The framework I’ll be presenting was described by media theorist Neil Postman in his book
Crazy Talk, Stupid Talk. It has long been out of print, but it’s highly readable and useful.
Many of the ideas in the remainder of this presentation come from Crazy Talk, Stupid Talk, so
if you’re interested in anything that I say from here on be sure to track it down.
Photo: http://bit.ly/1dc49Fn

Postman explains that traditionally we’ve thought of communication as something that
happens kind of like a ping-pong game: one person says something, the other listens,
considers a response, and replies, the ﬁrst person then listens, considers a response and
then replies, and so on. Postman suggests that communication is not like this at all.
Photo: http://www.ﬂickr.com/photos/theineffableme/3458421324/

Instead of something we do, communication is something we participate in, much like a plant
participates in what we call its growth. Growing is not something the plant does by itself; it
happens to it as it is exposed to nutrients, sunlight, and water. Growth is the result of the
plant’s exposure and interaction with its environment, and so it is with human
communication.
Photo: http://www.ﬂickr.com/photos/ralphhogaboom/297358105/

The semantic environment:
1. People (in all their glorious
complexity)
2. Their purposes in the situation
3. The rules of discourse by which such
purposes are achieved
4. The particular language that is
being used in the situation
So what are the earth, sun, and water of this environment that enables communication?
Postman lists four basic components: 1. The people who participate in the situation, 2. their
purposes, 3. the rules of discourse by which they achieve those purposes, and 4. the
language that is appropriate to the environment.

To illustrate, let’s look at two very different semantic environments: religion and science.
Each has its own goals, rules, and vocabulary. People participate in each for different reasons,
and the language that is appropriate in Church would get you tossed out of the research lab.
Photos:
http://www.ﬂickr.com/photos/catholicism/4091871705/
http://www.ﬂickr.com/photos/jeremywilburn/5229733212/

Semantic environments can also have sub-environments, with rules and vocabularies all their
own. For example, in the Catholic Church, there is a sub-environment called the
“confessional” which has its own speciﬁc rules and some technical vocabulary.
Photo: http://www.ﬂickr.com/photos/3059349393/3418857871/

Much trouble in the world is caused by people who do not understand what semantic
environment they’re actually participating in, as we can see from the constant clashes
between people who wish to enforce the goals, rules, and terminology of religion in the realm
of science, and vice-versa. Semantic environments are inherently different, and this
differentiation is important for us to be able to achieve our goals through them.

When this differentiation starts to be blurred — for example, when you try to use the rules
and terminology of religion within scientiﬁc discourse — we say that the semantic
environment is becoming polluted. We mean this in the same sense as pollution in the
physical environment. All environments contain some degree of foreign, or unassimilable,
matter in them. However, when there is too much unassimilable matter in the environment, it
becomes incapable of achieving its purposes, whether it is sustaining life or communicating
meaning.
Photo: http://en.wikipedia.org/wiki/File:AlfedPalmersmokestacks.jpg

Pay attention to the semantic
environment!*
* Hint: you are participating in one now.

The semantic environment:
• People
• Their purposes
• Rules of discourse
• Particular language
So let’s look more closely at the parts that make up the semantic environment. People are a
subject onto itself, and mainly out of our control, so I will not be talking much about them
today. Suffice it to say that we all parse situations through various ﬁlters which mold our
individual experiences and inﬂuence the way we understand and communicate with each
other.

Purposes
So let’s talk purposes, which is something we can actually understand and help support
through our work.

What we mean by purpose is best illustrated by example. Let’s think again about the
semantic environment we call “science”. You could say that one of the primary purposes of
science is to increase human knowledge through observation. The rules, relationships, and
vocabulary employed in this environment — for example, the process of peer review for
publication in journals like Science — support and enable this purpose.
Photo: http://en.wikipedia.org/wiki/File:Science_cover.gif

“Good” talk
There can be no meaningful concept of “good” or “bad” language without consideration for
the purpose of the environment. What is good in one environment is bad in another. For
example, the objective, unambiguous, detached, tentative, public language that is “good” for
a scientiﬁc environment would not be very effective during a romantic stroll on the beach.
Photo: http://www.ﬂickr.com/photos/absoblogginlutely/2858225616/

Remember: “good” talk does
what it’s supposed to do in a
particular situation!*
* As long as the purpose of the situation
is to serve rational and humane needs.

Contradicting
purposes
Most information environments have more than one purpose. Sometimes, these purposes are
in alignment. For example, the semantic environments employed by research scientists allows
them to advance their chosen fields and to advance their careers without conflict. However,
we often find environments where two or more purposes contradict each other.
Photo: http://www.flickr.com/photos/iita-media-library/4557461713/

For example, our mainstream news media have a hypothetical purpose of providing relevant,
reliable news coverage. However, most of them have an actual purpose of making money for
shareholders, primarily through advertising. Relevant, considered news is not the best
“eyeball magnet” for advertising, so many news outlets have devolved into semantic
environments that use the language and role-structures of news media, while bombarding us
with much content of questionable value.

Look out for conflict between
HYPOTHETICAL (or stated) versus
ACTUAL (or achieved) purposes!

Rules
Just like semantic environments have purposes, they also have special rules by which those
purposes can be achieved. Many of these rules have evolved out of cultural convention.

For example, not all situations can employ the same tone. It’s perfectly acceptable to whoop
and holler at a sports match, but this tone would be considered inappropriate in a wedding
ceremony.
Photo: http://www.ﬂickr.com/photos/amitchat/362493747/

Tú vs.
Usted
Another example, in Spanish we have two words that we can use to address people: tú, which
is the familiar you, and usted, which is more formal and honoriﬁc. Which you use depends on
who you’re talking to and in what context.

Understand the tone that is
most appropriate for the
environment!

A whole different set of rules has to do with role-structures: the relationships of power
between the people participating in the semantic environment. Think of your relationship
with airport security agents. Everything about the role-structure of this environment is
designed to disempower the passenger and empower the security agent. You can’t even joke
about it!
Photo: http://www.ﬂickr.com/photos/dcmetroblogger/8259826463/

Other semantic environments have much more fluid role-structures, for example, when you
hang out with your friends. There are no rules to which people are more sensitive than those
of role-structure. They are also very resistant to change. When there is a conflict between
authority and vocabulary, authority usually wins. Because of this, understanding the role-
structures of the environment is extremely important.
Photo: http://www.flickr.com/photos/soaress/7977994150/

Understand the role-
structures that influence the
environment!

Vocabulary
So we’ve talked about purposes and rules. Now lets talk about the words we use. Every
semantic environment has a set of words that have special meaning within that environment.
We call this a technical vocabulary.

“Responsive”
For example, in 2010 Ethan Marcotte introduced the word “responsive” into the semantic
environment of the web design community. It’s a term that helps us talk about important new
concepts — in other words, it serves a purpose — so it has caught on and become a part of
our technical vocabulary.

“Responsive” “Responsive”
Often, words may mean something in one environment and have a completely different
meaning in another. The meaning of “responsive” in the environment of the web design
community is completely different than its meaning in an environment like healthcare, where
it is also part of the technical vocabulary.

Every semantic environment
has a technical vocabulary.
Do you know yours?

When dealing with semantic environments, you also have to be on the lookout for what
Postman calls “key” words. These are simple, non-technical words which appear in every
semantic environment, but whose meanings shift as the words pass from one context to
another. Think for example of the word “law”. The fact that this word means something in the
Judeo-Christian tradition and something else when talking about the legal system causes a
great deal of trouble.
Photo: http://www.ﬂickr.com/photos/gerrydincher/6823441960/

“What is good for General Motors
is good for America.”
– Charles E. Wilson
Here’s another example, from a famous sentence you may have heard before: “What is good
for General Motors is good for America.” This presupposes that the word “good” in the ﬁrst
part of the sentence means the same as the word “good” in the second half. But obviously
this sentence is referring to two different semantic environments: that of business and that of
the nation-state, and the meaning of “good” is not the same in both.
Photo: http://history.defense.gov/wilson.shtml

Look out for “key” words:
seemingly simple words whose
meanings shift as they shift
contexts!

Metaphors
Finally, let’s talk about metaphors. Metaphors include both technical terms and key words,
but impose a particular imagery or point of view into the situation. For example, throughout
this presentation I’ve been asking you to think of communication as something that happens
in an “environment”, and have equated pollution in the physical environment with pollution in
this semantic environment. This is obviously a metaphor. While this imagery is useful to help
convey the concepts, we have to be wary of the fact that no metaphor is perfect and that our
thinking can become limited by the metaphor’s affordances.

Pay attention to metaphors!*
* Are you really “paying” attention?

Recap
The semantic environment consists of:
• People
• Their purposes
• Rules of discourse
• Particular language
To recap, the semantic environment consists of people, their purposes, the rules of discourse
by which such purposes are achieved, and the particular language that is being used in the
situation.

You’re going to be making
changes to this environment.
Don’t screw it up!

Craziness vs.
Stupidity
Well, Postman presents two pathologies: what he calls “crazy talk” and “stupid talk”. Let’s
look at the differences between them.

Stupid talk
Let’s start with stupid talk. Postman deﬁnes it as communication that has “(among other
difficulties) a confused direction or an inappropriate tone or a vocabulary not well-suited to
its context”. In other words, you’re saying the right words, but in the wrong tone, or in the
wrong place or situation. Because of this, it “does not and cannot achieve its purposes”.
Stupid talk usually has good intentions, but fails to deliver.

Think for example of a street-corner preacher. I’m sure you must have seen these folks
standing around in busy urban intersections, preaching salvation at the top of their lungs
while people just pass them by, pretending they aren’t there. The language and tone they’re
using may be perfectly effective when used in a church, but here it is out of sync with the
purposes of the other participants in the environment, so most people pass them by without
receiving the message.

Stupid is as
stupid does
When I say stupid, I don’t mean that the people involved are stupid. I mean that the semantic
environment itself is stupid. You can have very smart people on both sides, but the
environment makes it impossible for meaning to be transmitted effectively between them.
Let me give you another example, this one from my own career. I once did a heuristic
usability analysis of an online banking platform. In this system there was a button that was
very important: it was the button that took the user back to the “home” screen, which listed
all of their accounts. This button was present in every transactional screen.
Forrest Gump copyright © MCMXCIV by Paramount Pictures Corporation

Consolidated
The team that built the application used the word “consolidated” for the button’s label. This
term was part of the technical vocabulary used inside the bank. I had never seen it used
before in this context, and I suspected other people would have trouble with it too. I brought
the issue up with the team, and they were convinced that it wasn’t confusing. So I suggested
we test it with users. As you can imagine, nobody got it.

Consolidated
My accounts
We asked the users for a term that they thought would describe this functionality better, and
they suggested “my accounts”. Making this small change improved the usability of the
application considerably. The platform team had been forcing a term that was clear in one
semantic environment (that of bank IT) into another where it wasn't (that of the general
public). This was stupid: it was done with good intentions, but it didn’t work.

Crazy talk
Crazy talk is something else entirely. Again, quoting Postman, it is “talk that may be entirely
effective but which has unreasonable or evil, or sometimes, overwhelmingly trivial purposes.
It is talk that creates an irrational context for itself or sustains an irrational conception of
human interaction.” In other words, it is talk that works as intended, but has questionable
intensions. This is the territory of euphemisms and corporate double-speak.

For example, before coming out to San Diego I had to stop by a FedEx store to send some
documents to my office in Panama. Because they weren’t urgent (and I’m cheap), I selected
the package service labeled “FedEx Intl. Economy”. However, the FedEx agent helpfully
explained to me that “FedEx Intl. Priority” would actually be cheaper than “FedEx Intl.
Economy”. (He seemed a little baffled at the names too.) If I hadn’t been helped by an agent, I
would have overpaid by choosing “Economy”. This seems crazy to me.

The corporate world is full of sweet examples. Here is another one. Coca-Cola has a very nice
website that describes their social responsibility initiatives. It features large photos of
attractive people being physically active in natural surroundings.

This website has a section called “balanced living”, which includes information about calories
and obesity. I’ve heard there is an “obesity epidemic” in the U.S. and that sweet carbonated
drinks are one of the main culprits, so I was curious about what Coca-Cola had to say on the
matter.

This is where I learned that I’m getting fat from eating too much chicken. I thought, “this
can’t be, where are the sodas?” So I decided to track down the source of this information.
(Which isn’t hard to do: as you can see, Coca-Cola has helpfully included a footnote with the
reference.) This particular bit of information comes from the 2010 Dietary Guidelines
for Americans, published by the USDA.

So I downloaded this document, and found out that — sure enough — the top three culprits
are cakes, breads, and chicken dishes. But guess what number four is? I don’t think it’s a
coincidence that the Coca-Cola website presents only the top three, and that it does so in a
context that appears to be educational, transparent, and neutral. This seems crazy.

Stupid talk
Crazy talk
So I’ll summarize the difference between them by again quoting Postman: “Talk is stupid
when it does not work. Talk is crazy when, in working, it creates and sustains an irrational
purpose.”
Fixing stupid talk is our professional responsibility. Fixing crazy talk is our ethical
responsibility.

Your job is to fix craziness or
stupidity when you
encounter it!*
* Or get out while you can!

… in practice
So now that we’ve seen the theory, I’m going to show an example of you how you can take
these ideas out into the ﬁeld and use them to help make better IA decisions. This is a
technique my team and I have used at the beginning of projects to map the semantic
environments that make it up.

Steps
1. Workshops with clients and users
2. Figure out how many semantic
environments we’re dealing with
3. Have both parties map these
environments
4. Consolidate the maps
At the beginning of the project, we ask the client to set up workshop-style meetings with
different groups of people who interact in their processes. This includes internal people as
well as their clients. In these workshops, we try to identify how many semantic environments
we’re dealing with, and have both parties map them. Then we consolidate them.

Typical questions
• Who participates in this process?
• How do you interact with them?
• What outcome(s) do you expect?
• What questions do you ask?
• What words are effective?
• What words turn them off?
• Are there any gatekeepers?
We start the workshops by explaining to participants the concept of the semantic
environment, and the nature of the exercise. Then, as a group, we ask them questions such
as these. We note their answers in sticky notes, which at ﬁrst don’t have much structure.
Eventually, patterns emerge. We’re looking for groups of sticky notes that seem to deﬁne
individual semantic environments.

We then set up a ﬂip chart for each different semantic environment we identify, and start
moving the smaller sticky notes to them in a particular structure:

Actors Rules
Goals Words
We have one quadrant for each of the four components of the semantic environment: actors,
or the people who participate in the environment, goals, rules, and the terms they use.

Actors Rules
Goals Words
PARTYA
PARTYB
We subdivide the actors/goals half of the grid into two columns, one for the ﬁrst party, in this
case our client, and the other for the second party, in this case, their clients. This is because
both parties have different people and purposes in these environments, and we want to
understand their perception of what the other party will want. Note that we will be repeating
the exercise with the second party, so we get both perspectives.

The people involved →
Their goals →
The rules →
The vocabulary →
Interaction channels →
When these mapping exercises are done, we consolidate the responses into individual maps
that include the perspective of both parties. Note that here we also have two columns for the
rules and vocabulary. This is because we want to identify terminology and rules that they
consider positive and negative; that help or hinder both parties to achieve their purposes.
We’ve also added a row to document both current and possible interaction channels.

As I’ve hinted, we are mapping multiple semantic environments and sub-environments. We
want to get as speciﬁc as possible. For example, in this case, we had a semantic environment
at the beginning of a relationship with a prospect, when executives are involved from both
parties. We also had a separate environment that became active once the parties decided to
enter into a formal business relationship, at which point lawyers and ﬁnancial teams take
over. They all have different objectives, role-structures, tones, and vocabularies, so we want
to map them separately.

“Monitor” = Good!
“Monitor” = Bad!
Seeing these maps side-by-side helps you see if the technical vocabulary and rules match up
for both parties across the board, and if there are any gaps you need to ﬁll in. You can also
spot roadblocks and opportunities inherent in language. For example, in the course of this
exercise, we identiﬁed the word “monitor” as a potential key word, that had positive
meanings from the perspective of our client but meant something else, with negative
connotations, from the perspective of their clients.

It’s worth noting that there aren’t hard boundaries between these semantic environments —
they inform and inﬂuence each other, sometimes in a sequence. Also, these maps are never
fully complete. The value in the process comes from raising the questions and identifying
unmet challenges and opportunities in language.

What we’ve learned:
1. Understanding happens in semantic
environments
2. These environments can be polluted,
turning stupid or crazy
3. You can help reduce the stupidity and
craziness in the world
4. The first step is to map the semantic
environments you’ll be working in

Map before you yap!

Coda
I will close with one final thought. I believe that understanding and shaping semantic
environments are critical steps in the development of an information architecture. These
things are central to what we do. I think we’re doing much of it without calling it such, and
many of you are doing it well. However, I believe we’re not doing a very good job of
understanding and influencing the semantic environments that pertain to how we explain this
to our clients, our peers in other disciplines, and the world at large. I think that by applying
some of these ideas onto our own field, we can make headway.

Let me give you an example of what I mean. Just last week I had the opportunity to see a live
debate between two people I admire tremendously: Dan Klyn, who is here with us today, and
Matt Nish-Lapidus from the IxDA. Each of them did a very good job of framing the disciplines
of IA and Interaction Design respectively, and arguing the merits of their views. I got a great
deal of value from this discussion, and I only bring it up here because it reminded me that as
far as I recall, we’ve been thinking of these two areas of focus in a sort of parallel role.

This view goes a ways back, and was reinforced in Jesse James Garret’s milestone closing
plenary at the IA Summit in Memphis, where he spoke of the two professions in a way that
makes them sound like sister disciplines: “There are no information architects. There are no
interaction designers. There are only, and only ever have been, user experience designers.”

Information
Architecture
User Experience Design
Interaction
Design
This suggests to me that the abstraction model that many of us have in our minds looks
something like this, where both disciplines are areas of focus within a broader ﬁeld called UX.
In my mind, this model has serious limitations. For one, I don’t believe that you can show an
example of “pure” information architecture in the world, devoid of instantiation, any more
than you can photograph a “pure” musician. And every time I see something made by an
interaction designer, I see information architecture in it.

Information
Architecture
Interaction
Design
Composition
Guitar-playing
ABSTRACTION
I’ve increasingly come to think that a more productive model for this relationship would be to
think of interaction design as happening in a totally different level of abstraction from
information architecture, in much the same way that you can say that guitar-playing is in a
different level of abstraction from the composition you’re hearing. You wouldn’t be able to
experience music without people to play it; you would only be able to appreciate it
intellectually, which is not the same thing. Note also that guitar-playing is not the only way to
get your composition heard.
Of course, not everyone who plays guitar will be playing complicated pieces for which they
need sheet music. But whether you’re playing Sibelius or the Sex Pistols, a little knowledge of
composition goes a long way to making better music. Note that this does not mean that one
is more important than the other. It also doesn’t mean that they’re two sides of the same
coin. It means that they are intertwined in a symbiotic semantic relationship at different levels
of abstraction, and this inﬂuences the way we communicate. Understanding this could
improve our ability to work together, as we must, since we can’t experience one without the
other.

Give a hoot! Mapping (and caring for) the semantic environment

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Give a hoot! Mapping (and caring for) the semantic environment (20)

More from Jorge Arango (15)

Recently uploaded (20)

Give a hoot! Mapping (and caring for) the semantic environment