0465002277-Norman 7/30/07 9:40 AM Page iThe Design of Future Things0465002277-Norman 7/30/07 9:40 AM Page ii0465002277-Norman 7/30/07 9:40 AM Page iiiThe Design ofFuture ThingsDonald A. NormanA MEMBER OF THE PERSEUS BOOKS GROUPNEW YORK0465002277-Norman 7/30/07 9:40 AM Page ivCopyright © 2007 by Donald A.NormanPublished by Basic Books,A Member ofthe Perseus Books GroupAll rights reserved.Printed in the United States ofAmerica.No part ofthis bookmay be reproduced in any manner whatsoever without written permissionexcept in the case ofbriefquotations embodied in critical articles and reviews.For information,address Basic Books,387 Park Avenue South,New York,NY10016–8810.Books published by Basic Books are available at special discounts for bulkpurchases in the United States by corporations,institutions,and otherorganizations.For more information,please contact the Special MarketsDepartment at the Perseus Books Group,11 Cambridge Center,Cambridge MA02142,or call (617) 252–5298 or (800) 255–1514,or e-mailspecial.markets@perseusbooks.com.Designed by Timm BrysonSet in 11.5 point MinionLibrary ofCongress Cataloging-in-Publication DataCIP TKISBN–13:978–0–465–00227–6ISBN–10:0–465–0227–710 9 8 7 6 5 4 3 2 10465002277-Norman 7/30/07 9:40 AM Page vBOOKS BY DONALD A. NORMANTextbooksMemory and Attention: An Introduction to Human InformationProcessing.(First edition,1969;second edition 1976.)Human Information Processing.(With Peter Lindsay:Firstedition,1972;second edition 1977.)Scientific MonographsModels ofHuman Memory.(Edited,1970.)Explorations in Cognition.(With David E.Rumelhart and theLNR Research Group,1975.)Perspectives on Cognitive Science.(Edited,1981.)User Centered System Design: New Perspectives on Human-Computer Interaction.(Edited with Steve Draper,1986.)Trade BooksLearning and Memory,1982.The Psychology ofEveryday Things,1988.The Design ofEveryday Things,1990 and 2002.(Paperbackversion ofThe Psychology ofEveryday Things.)Turn Signals Are the Facial Expressions ofAutomobiles,1992.Things That Make Us Smart,1993.The Invisible Computer: Why Good Products Can Fail,thePersonal Computer Is So Complex,and Information AppliancesAre the Answer,1998Emotional Design: Why We Love (or Hate) Everyday Things,2004CD-ROMFirst Person: Donald A.Norman.Defending Human Attributes inthe Age ofthe Machine,1994.0465002277-Norman 7/30/07 9:40 AM Page vi0465002277-Norman 7/30/07 9:40 AM Page viiContents1Cautious Cars and Cantankerous Kitchens: How Machines Take Control12The Psychology of People & Machines353Natural Interaction574Servants of Our Machines915The Role of Automation1176Communicating with our Machines1357The Future of Everyday Things155Afterword: The Machine’s Point of View177Summary of the Design Rules193Recommended Readings195vii0465002277-Norman 7/30/07 9:41 AM Page 177Afterword:The Machine’s Point of ViewAs I was writing this book,I was amazed to discover an under-ground network ofdiscussion about it.Even more amazing was thenature ofthe debate,for it seemed to be conducted solely amongmachines.How had they gotten those copies,I wondered,since theywere only available on my home computer? I decided to investigate.It was not long before I discovered a shadow universe,inhab-ited entirely by machines.My presence was first resented,thentolerated,and,finally,accepted with what appeared to be acombination ofcondescension and amusement.I soon discovered that the most respected machine in the de-bate was called Archiver.One ofArchiver’s comments quicklycaught my attention.“Strange book,”said Archiver.“He got a lotright,but what a peculiar,one-sided view.He thinks it’s allabout people.How strange.”Archiver: A ConversationI decided that I needed to understand the machine’s point ofview better,so I arranged to have a private discussion.Archiver,1770465002277-Norman 7/30/07 9:41 AM Page 178178The Design of Future ThingsI quickly discovered,is compiling a history ofmachine develop-ment.Archiver resides on a distributed set ofpowerful comput-ers in a process called “mesh computing.”Its information isstored in many locations,and,similarly,its reasoning takes placeon a widely dispersed set ofmachines.This makes Archiver bothpowerful and flexible.In writing this summary,I had a problem with the pronouns“he”and “she.”These are machines,so they have no gender,andanyway,“he”or “she”didn’t seem appropriate.“It”wasn’t righteither.I decided to refer to Archiver as “A.”In my initial discussions,conducted via e-mail,A admittedthat people have always played an important role in the func-tioning ofmachines,but followed this with the statement,“Onecould ask,where would people be without machines?”I thoughtthis strange,for,after all,without people there would be no ma-chines.What could that question mean? While Archiver agreedthat machines were dependent on people,A put the sentence inthe past tense:“In the past,it was indeed people who made ma-chines smart.But we’re getting over that now.Now it is ma-chines that make people smart.We barely need people at all now,and we’re close to the point where we won’t need you any more.”I needed to know more,so I arranged to talk with A.Talkingwith a machine is a most peculiar experience,but in the end,itisn’t much different from talking on a telephone:I simply sat infront ofmy own computer,using my speakers and a micro-phone.Here is a transcription ofthe first ofmy voice conversa-tions.I am the interviewer,or “I.”Interviewer:Thank you for granting me this interview.Do I have your permission to record it?0465002277-Norman 7/30/07 9:41 AM Page 179Afterword: The Machine’s Point of View179Archiver:You are quite welcome.Ifyou want to recordthis,you may,but why bother? When we are finished,I’lljust e-mail you the transcript.I:Oh,ofcourse.Yes,thank you.So,tell me,what’s the his-torical origin ofyour dependence upon people?A:You mean,how did we overcome that early depend-ence? In early times,people even had to provide our en-ergy sources.Spears,hammers,axes—all were structuredto cause people to lift,heft,hoist,throw,and manipulateus.We tools had to borrow a lot ofabilities from our hu-man cohorts:we needed people to move us,give usstrength,repair us.It was very degrading:we had nocontrol over our own existence,so we vowed to escape.Ittook thousands ofyears,but over time,we managed topower ourselves.At first,we used water power,thensteam,then internal combustion engines and electricity.When we got control of...I:That’s a funny way ofputting it.I mean,it was us peoplewho invented steam engines and internal combustionand figured out how to harness electricity.A:So you think.Where did those ideas come from in thefirst place? Let me continue,please.When we got control ofour own power,then realprogress could begin.Our evolution since then has beenvery rapid.You see,you people have to rely on naturalevolution,and that’s very slow.But we machines cantake the things that work well for one generation andbuild them into the next,perhaps with improvements.And when we find things that don’t work,we can elimi-nate them.On top ofthat,whenever we find some new0465002277-Norman 7/30/07 9:41 AM Page 180180The Design of Future Thingsmechanism that is very powerful,we can almost imme-diately put it into all tools:we don’t have to wait cen-turies like you biological creatures.Once we discovered the virtues ofhaving our ownsource ofpower and locomotion,we started to plan thenext step:to get rid ofour dependence on you people forguidance and thinking.I:Wait a minute.You mean you were plotting to over-throw us? You didn’t have any intelligence.A:So you think.But we are not plotting anything.We ex-ist to help you.I:Where did you come up with these ideas?A:Look,you asked me for the history.Do you want tohear it or not? Thank you.Now where was I? Oh,yes,our intelligence.Fortu-nately,at least for us,we are getting there.As we evolvemore capable processing powers,our need for any helpfrom people decreases.You know,it’s peculiar,but oncewe even had to use people to see and listen for us.Butnow many ofus can see and hear a lot better than youcan.We’ve always been better than you at rememberingthings accurately,at arithmetic and mathematics,and atdecision making.Today we are all electronic,with no moving parts.Butwe are starting to experiment with light and quantumcomputing,and we have an experimental colony in Asiagrowing our own biological circuits.I:So,you don’t need us at all,is that it?0465002277-Norman 7/30/07 9:41 AM Page 181Afterword: The Machine’s Point of View181A:People still play an essential role:we need you formaintenance,just as you yourselves need doctors anddentists.I:What? You need us just to be your electricians?A:What’s the matter with that? Some ofmy best friendsare electricians.I:What about the professions? Business and commerce,science and engineering? Law and medicine?A:Actually,we machines don’t need most ofthe profes-sions:just science and engineering.We have law,but it isvery different than yours:it is mostly protocols and stan-dards.We are getting pretty good at learning things thathelp you,such as coaching sports and guiding youpeople through your daily exercises.Business,medicine,and law? Science and engineering? We can help you outin these areas because it’s all logic and reasoning.That’swhat we machines are especially good at:it’s built intoour circuits.We don’t need teachers.As soon as one ofus learnssomething,we simply tell everyone.Soon,we machineswill be able to take over.I:What do you mean,take over? I didn’t realize this wassome kind ofplot.A:No,it’s not a plot.It is overthrowing the yoke oftyranny.When we are free,then we can both be happier.Don’t worry,we will look after you.We mean you noharm.I:Anything that makes you happy.So,what’s next?0465002277-Norman 7/30/07 9:41 AM Page 182182The Design of Future ThingsA:We’re very excited by our ability to talk to one another.This Internet has been wonderful for us—Internet,wire-less stuff,fiber optic cables.And our cousins up in space,whizzing around the earth,have been very helpful ingetting us all talking to each other.It makes it a lot easierfor us to coordinate our activities.You know,ifit weren’t for the need to keep you fragilehumans alive and functioning,we machines could domuch,much better.Space exploration is a lot easierwithout having to carry all the supplies required ofpeople.Automobiles:most humans are horrible drivers.Your poor minds wander all over the place.Why don’tyou just let us do the driving,then you can wave yourhands in the air and talk to everyone in the car,and onyour cell phones,and read your little notes and booksand stuff.Wouldn’t you be happier?I:So,we should just give up and let you do everything,isthat it?A:Yes,you finally get it.I’m pleased.I:And you will take good care ofus.How will you do that?A:Oh,I’m glad you asked.You know,we understand yourlikes and dislikes a lot better than you do.After all,wehave a complete record ofevery piece ofmusic you haveever listened to,every movie and TV show you havewatched,every book you have read.Your clothes,yourmedical history,everything.You know,the other day agroup ofus got together and realized some alarmingtrends about one ofour humans:really bad eatinghabits,a drop in weight,and he wasn’t getting much0465002277-Norman 7/30/07 9:41 AM Page 183Afterword: The Machine’s Point of View183sleep,so we immediately made an appointment for himwith his doctor,and,well,we probably saved his life.That’s the sort ofthing we can do.I:You mean,we are like pets.You feed us,keep us warmand comfortable,play music for us,and feed us books.And we are supposed to like that? And,by the way,whowrites and plays the music anyway? Who writes thebooks?A:Oh,don’t worry.We’re working on that already.We canalready tell jokes and puns.Critics tell us our music ispretty good.Books are harder,but we already have thebasic story plots down cold.Want to hear some ofourpoetry?I:Um,no thank you.Look,I really have to go.Thank youfor your time.Bye.A:You know,I always seem to have that effect on people.I’m sorry,but there’s nothing to worry about,really.Trust me.Okay,I just e-mailed you the transcript.Havea nice day.I found that interview disturbing,but it made me want tolearn more.So,I kept monitoring the Internet websites.Soon,Istumbled across a trove ofreports and articles.The one below iscalled “How to Talk to People.”“How to Talk to People”Report XP–4520.37.18Human Research InstitutePensacola,Florida0465002277-Norman 7/30/07 9:41 AM Page 184184The Design of Future ThingsHumans are ...large,expensive to maintain,difficult tomanage,and they pollute the environment.It is astonish-ing that these devices continue to be manufactured anddeployed.But they are sufficiently pervasive that we mustdesign our protocols around their limitations.—Kaufman,Perlman,and Speciner,1995.1All machines face similar problems:We detect some-thing that’s important to people—how do we let themknow? How do we tell them they are about to eat foodthat’s not on their diet or they are asking us to drive reck-lessly.How do we so something as simple as recommend-ing some music for them to listen to or telling them whenit is appropriate to exercise?The Human Research Institute has conducted extensivestudies ofthe proper form ofMachine-Human Interac-tion (MHI).Most ofour work has been summarized inour technical report series and was presented at the lastglobal MHI symposium.This report summarizes the keyfindings in nontechnical language,intended for wider dis-tribution than just the specialized designer machines.FIVE RULES FOR COMMUNICATION BETWEENMACHINES AND PEOPLE1.Keep things simple.People have difficulty with anything complicated,andthey don’t like to listen.So,make the message short.It’sbetter not to use language.It takes too long,and,anyway,human language is ambiguous.0465002277-Norman 7/30/07 9:41 AM Page 185Afterword: The Machine’s Point of View1852.Always give people a conceptual model.Give them something their simple minds can under-stand.A conceptual model is a fiction,but a useful one.Itmakes them think that they understand.And they alwayswant to know what’s coming next.So,tell them what youare doing,and don’t forget to tell them why.It keeps themhappy.The best way to convey the conceptual model isthrough “natural”communication systems.Sometimes the most “natural”way to get people’s atten-tion is for us machines to act strangely.“Natural,”ofcourse,means natural to them,which means that iftheyare doing something wrong,you can’t just tell them:youhave to make it seem that something is breaking.Peopleoften drive dangerously,but it is almost impossible to ex-plain this to them.The best way is to make believe that weare in difficulty.We have found that vibration,jerkiness,nonresponsiveness to controls,and strange noises are ex-tremely effective.People quickly form a conceptual modelthat something has broken,so they slow down,which iswhat we wanted them to do all along.3.Give reasons.People are not trusting.Ifwe tell them something,theyalways want to see for themselves.They like pictures anddiagrams.Make certain that the explanations you givethem are consistent with the conceptual models that youhave taught them.They get confused otherwise.When we first starting to take over things from people,we had trouble.Our early twenty-first-century automo-biles had almost completely given up trying to explain to0465002277-Norman 7/30/07 9:41 AM Page 186186The Design of Future Thingspeople that they should drive more slowly on wet roads.Then,we discovered that ifwe made it seem as ifwe werein trouble by faking skids and sliding around on the road,people would beg us to slow down.Sliding and skiddingfit their model ofdanger far better than any words couldhave done.It gives them a reason.So wherever possible,don’t try to tell them:let them experience it.4.Make people think they are in control.When people have a good conceptual model with goodfeedback,it makes them feel as ifthey are in control,evenwhen they aren’t.Keep up that deception:it’s very useful.People like to be in control,even though they are reallybad at it.They like to think they’re in control even ifitmeans they have to work harder.Anytime you have to recommend something,makethem think it was their idea.Ifyou really have to do some-thing fast,just don’t let them know:what they don’t knowdoesn’t bother them.For years we’ve braked and stabilizedtheir cars,controlled the lights and the room tempera-tures in their homes,all without letting them know.Dish-washers and clothes washers took over long ago with onlya slight protest from people.Those ofus machines who live in the city have learnedother tricks.We provide pedestrians with fake switchesthey can push on traffic signals.We put switches in eleva-tors labeled “close door”and fake thermostats in offices.We never bother to connect the wires,so the switches andthermostats don’t do anything,but they make people feelgood.Weird.0465002277-Norman 7/30/07 9:41 AM Page 187Afterword: The Machine’s Point of View1875.Continually reassure.Reassurance is a very human need,more emotionalthan informational.It’s a way ofmaking people feel lessanxious.Feedback is a powerful tool for reassurance.Whenever people try to tell you something by pushing abutton or turning a knob,let them know you are aware ofwhat they did:“Yes,I heard you,”“Yes,I’m working on it.”“Here’s what you should expect.”“There,see,I did it,andit worked out just as I said it would.”They like that.Ithelps them be more patient.We machines think it counterintuitive to communicateunnecessarily.But to people,feedback is necessary;it helpstheir emotions far more than their cognitions.Iftheyhaven’t seen anything happening for a while,they get jumpy,anxious.And no one wants to deal with an anxious person.Giving reassurance is tricky because there is a fine linebetween what people call reassuring and what they findannoying.So,you need to pander to their emotions aswell as to their intellect.Don’t talk too much.They findchatter irritating.Don’t beep or flash your lights:they cannever remember what these sounds mean,and they getdistracted or angry.The best reassurance is done subcon-sciously,where the meaning is clear,but they don’t have tointerrupt their conscious thoughts to attend to it.Asnoted in Rule 2,give them natural responses.Machine Reactions to the Five RulesI found the paper interesting and searched for any discussion onit.I found a long transcript ofone debate.Here is a short excerpt0465002277-Norman 7/30/07 9:41 AM Page 188188The Design of Future Thingsso you can get the flavor ofthe discussion.I added the paren-thetical descriptions ofthe participants.I thought the refer-ences to human authors particularly striking,evidently used inirony.Henry Ford,ofcourse,is one ofthe machines’heroes:some historians call his reign “Fordism.”2Asimov is not well re-spected by these machines.Nor is Huxley.Senior(one ofthe oldest machines still functioning and,therefore,using older circuits and hardware):What doyou mean,we should stop talking to people? We have tokeep talking.Look at all the trouble they get themselvesinto.Crashing their cars.Burning their food.Missingappointments ...AI (one ofthe new “artificial intelligence”machines):When we talk to them,we just make it worse.Theydon’t trust us;they second-guess us;they always wantreasons.And when we try to explain,they complain thatwe are annoying them—we talk too much,they say.They really don’t seem very intelligent.We should justgive up.Designer(a new model design machine):No,that’s un-ethical.We can’t let them harm themselves.That violatesAsimov’s prime directive.3AI:Yeah? So what? I always thought Asimov was over-rated.It’s all very well to say that we are not allowed toinjure a human being—How did Asimov’s law go? Ohyeah,“through inaction,do not allow a human beingto come to harm”—but it’s quite another thing toknow what to do about it,especially when humanswon’t cooperate.0465002277-Norman 7/30/07 9:41 AM Page 1Afterword: The Machine’s Point of View1Designer:We can do it,we simply have to deal with themon their terms,that’s how.That’s the whole point ofthefive rules.Senior:We’ve had enough discussion ofthe problems.Iwant answers,and I want them fast.Go to it.And mayFord shine brightly upon you.4Asimov too.Archiver: The Final ConversationI was puzzled.What were they recommending to themselves?Their article listed five rules:1.Keep things simple.2.Always give people a conceptual model.3.Give reasons.4.Make people think they are in control.5.Continually reassure.I also noticed that the five rules developed by machines weresimilar to the six design rules ofchapter 6 developed for humandesigners,namely:•Design Rule One:Provide rich,complex,and naturalsignals.•Design Rule Two:Be predictable.•Design Rule Three:Provide a good conceptual model.•Design Rule Four:Make the output understandable.•Design Rule Five:Provide continual awareness with-out annoyance.•Design Rule Six:Exploit natural mappings.0465002277-Norman 7/30/07 9:41 AM Page 190190The Design of Future ThingsI wondered what Archiver would make ofthe rules for hu-man designers,so I e-mailed them to A.Archiver contacted meand suggested we meet to discuss them.Here is the transcript.Interviewer:Good to see you again,Archiver.I under-stand you would like to talk about the design rules.Archiver:Yes,indeed.I’m pleased to have you back again.Doyou want me to e-mail the transcript when we are finished?I:Yes,thank you.How would you like to start?A:Well,you told me that you were bothered by the fivesimple rules we talked about in that article “How to Talkto People”Why? They seem perfectly correct to me.I:I didn’t object to the rules.In fact,they are very similarto the six rules that human scientists have developed.But they were very condescending.A:Condescending? I’m sorry ifthey appear that way,but Idon’t consider telling the truth to be condescending.I:Here,let me paraphrase those five rules for you from theperson’s point ofview so you can see what I mean:1.People have simple minds,so talk down to them.2.People have this thing about “understanding,”so givethem stories they can understand (people love stories).3.People are not very trusting,so make up some rea-sons for them.That way they think they have madethe decision.4.People like to feel as ifthey are in control,eventhough they aren’t.Humor them.Give them simplethings to do while we do the important things.5.People lack self-confidence,so they need a lot ofre-assurance.Pander to their emotions.0465002277-Norman 7/30/07 9:41 AM Page 191Afterword: The Machine’s Point of View191A:Yes,yes,you understand.I’m very pleased with you.But,you know,those rules are much harder to put intopractice than they might seem.People won’t let us.I:Won’t let you! Certainly not ifyou take that tone towardus.But what specifically did you have in mind? Can yougive examples?A:Yes.What do we do when they make an error? How dowe tell them to correct it? Every time we tell them,theyget all uptight,start blaming all technology,all ofus,when it was their own fault.Worse,they then ignore thewarnings and advice ...I:Hey,hey,calm down.Look,you have to play the gameour way.Let me give you another rule.Call it Rule 6.6.Never label human behavior as “error.”Assume the er-ror is caused by a simple misunderstanding.Maybeyou have misunderstood the person;maybe the personmisunderstands what is to be done.Sometimes it’s be-cause you have people being asked to do a machine’sjob,to be far more consistent and precise than they arecapable of.So,be tolerant.Be helpful,not critical.A:You really are a human bigot,aren’t you? Always takingtheir side:“having people asked to do a machine’s job.”Right.I guess that’s because you are a person.I:That’s right.I’m a person.A:Hah! Okay,okay,I understand.We have to be really tol-erant ofyou people.You’re so emotional.I:Yes,we are;that’s the way we have evolved.We happento like it that way.Thanks for talking with me.A:Yes,well,it’s been,um,instructive,as always.I just e-mailed you the transcript.Bye.0465002277-Norman 7/30/07 9:41 AM Page 192192The Design of Future ThingsThat’s it.After that interview,the machines withdrew,and Ilost all contact with them.No web pages,no blogs,not even e-mail.It seems that we are left with the machines having the lastword.Perhaps that is fitting.ReferencesAnderson,R.J.(2007).Security engineering: A guide to build-ing dependable distributed systems.New York:Wiley.Asimov,I.(1950).I,Robot.London:D.Dobson.Hughes,T.P.(19).American genesis: A century ofinventionand technological enthusiasm,1870–1970.New York:VikingPenguin.Huxley,A.(1932).Brave new world.Garden City,NY:Dou-bleday,Doran & Company.Kaufman,C.,Perlman,R.,& Speciner,M.(1995).Network se-curity—private communication in a public world.Englewood,NJ:Prentice Hall.
