I will admit, I was entirely biased upon reading this story from Nick Carr, who has a knack for writing pieces that get a lot of attention by baiting his hook with contrarian link chum. Heck, he’s really good at it, and I have a lot of respect for Nick. So I’ll take the bait.
His piece starts by conjuring HAL, the famous AI which manipulates humans, then makes his case by citing his own “feeling” that Google has changed his attention span to somehow prove that search and web browsing in general is making us stupid.
Balderdash. What Carr is really saying is this: People are not reading long narrative anymore, and that makes me and my pals sad. So let’s blame the Internet!
Sounds an awful lot like the complaints we heard about TV making us stupid. Did TV make us stupid? I dunno, ask Steven Johnson. I bet he has an opinion on this piece as well.
Carr writes: “Yet, for all that’s been written about the Net, there’s been little consideration of how, exactly, it’s reprogramming us. The Net’s intellectual ethic remains obscure.”
So because Nick hasn’t come up with a singular thesis as to what the “Net’s intellectual ethic” is, we must declare it’s making us stupid, eh?
Huh. He goes on to claim that Google is, in essence, an industrial style factory driven by a philosophy that is mechanizing our collective intellect much like factory automation mechanized our collective workforce – in short, Google is turn our minds into nothing more than collective cogs in some borg like hive mind. We’re fucked, and it’s all Google’s fault.
Puuuuuuuhhhhleezzze.
Here’s another quote: “The last thing these companies want is to encourage leisurely reading or slow, concentrated thought. It’s in their economic interest to drive us to distraction.”
Right. And that’s why Google encourages its workers to spend 20% of their time on passion projects. OK.
His conclusion: “As we come to rely on computers to mediate our understanding of the world, it is our own intelligence that flattens into artificial intelligence.”
Good lord. Somehow Carr seems to presume that there’s simply nothing valuable occurring in our minds when we engage with the extraordinary new medium of the web. Because we’re starting to think in different ways, it must be bad. Right? Carr may believe that search and the Internet make us stupid, but I will counter his personal, anecdote-driven conclusions with one of my own: when I am deep in search for knowledge on the web, jumping from link to link, reading deeply in one moment, skimming hundreds of links the next, when I am pulling back to formulate and reformulate queries and devouring new connections as quickly as Google and the Web can serve them up, when I am performing bricolage in real time over the course of hours, I am “feeling” my brain light up, I and “feeling” like I’m getting smarter. A lot smarter, and in a way that only a human can be smarter.
And I have a feeling I’m not alone. What do you guys think?
I would agree with you. How the pursuit of knowledge can lead to the suffocation and dampening of learning is something that I can’t come to terms with nor agree with. Search innovation has made information more easily accessible for users. Some may argue that this ease of access and publication has led to a decline in the quality of content we encounter. I would argue that there is value in the “declined” value because it forces us, as learners, to discern what we consider the “truth”. Back in the day we would read an encyclopedia and consider it absolute. If you ask me, Search information and Google have pushed us away from the “intelligence flattening” of the dictionary and encyclopedia days. You’ve got my support, John!
Thank you.
As much as I admire his ability to see emerging trends in their composite, Carr’s fear-mongering conservatism is often much too one-sided.
Great post, John.
“contrarian link chum” = best description I’ve heard to date. 🙂
I’ll admit, sometimes I find myself with the laptop open in the middle of the evening, and I ask myself: What have I been doing for the last hour or two? It often amounts to relatively low-level grazing. I’m sure that’s my fault, and I don’t blame the Internet. But if I don’t watch it, I follow the path of least resistance. This is harder to do when you’ve made a commitment to read a book. That said, a lot of people aren’t replacing the Internet for books, but for TV.
not alone. i didn’t even read it. just the title tells you it’s a lot of pontification and handwaving.
Both these posts are awesome! I think Google is doing a great job but it is still far far away from being anywhere close to HAL metaphor! A knowledge worker still has to spend inordinate amounts of time researching (http://searchengineland.com/071023-093541.php) but this should make him smarter, not stupider!
In the words of Herbert Simon, “What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention, and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.”
An average person is getting access to so much of information for the first time, and I guess this is making some unforeseen repercussions in the way we function. We still have to learn to cope with so much of information and I believe there is a major role of technology and collaboration here!
Thanks – very insightful points of view! 🙂
Wasn’t Socrates against writing as it took away memory from the mind and made it less nimble? Good thing Plato wrote down “the dialogues” or we would know so much about Socrates. Writing has seamed to work out OK so far – I hope the Internet can do as well.
The article was poorly named. But very interesting and thought provoking. I am convinced that the way we think does change with long-term use of the web but have not decided whether this is good or bad. The change is compelling and feels good but is there a difference in quality?
Well, unsurprisingly, I tend to agree.
By only providing us with 10 links, ads on the right, and then an occasional spell check, an engine like Google does a terrible job of providing a contextual, explanatory, “sensemaking” overview of the 1.2 million hits that come back as a result of one of my information-seeking queries.
Thus, I think there is some merit to the point that Google makes us dumber, because it conditions us to believe that its top 10 results are (1) the best answers, and (2) the only answers. Even if it says that there are 1.2 million hits, Google gives us no informational, sensemaking instruments to help us deal with those 1.2 million hits, and thereby encourages us to lazily not bother. Who can possibly scroll through 1.2 million results, 10 links at a time? And even if you do have the patience to manually slog through, say 500 of those 1.2 million results, how does Google really help you organize and categorize and sort through what you’re finding, as you find it?
What I would really like to see is something like Vivisimo, that gives you a clustered overview of the entire range of possible different results to your query. So when you query for “causes of poverty” it offers to you a sensical overview of the topic space, i.e.: community, issues, hunger, research, solutions, families, social, reduction, children, etc.
http://clusty.com/search?input-form=clusty-simple&v%3Asources=webplus&query=causes+of+poverty
Or Ask, that gives you helpful ways of both narrowing and expanding your search, which is also another form of information visualization and discovery, something that helps me think a little bit deeper about the topic at hand:
http://www.ask.com/web?q=causes+of+poverty&search=search&qsrc=0&o=0&l=dir
[Disclaimer: I do not work for either of those companies]
Such information helps you open up your vision, and think more, and discover aspects of your topic that might not have occurred to you, otherwise. Vivisimo and Ask encourage smarter, lateral thinking, whereas Google only provides ads.
So I tend to agree when Carr says: “The last thing these companies want is to encourage leisurely reading or slow, concentrated thought. It’s in their economic interest to drive us to distraction.”
Google thinks that there is more user utility to taking up screen space by showing an ad than there is to dynamically contextualizing your SERPs, helping you understand the conceptual space returned by your information-seeking query a little bit more.
It certainly is in their economic interest to make sure we’re driven to distraction, rather than consciously, deeply engaged with the data. They prove it by their actions, by the choice of information they show as a result of user queries.
And John, I don’t understand why your counterargument to Nick’s contention is to cite the 20% rule:
Right. And that’s why Google encourages its workers to spend 20% of their time on passion projects. OK.
What does the 20% time have to do with economic interest in driving us to distraction?
If anything, given the amount of non-advertising-based money that all these 20% projects have added to the Google bottom line (which is to say, um, almost none — how much money did Google make on Google Web Accelerator, for example), I’d say that 20% has zero bearing on what Google does in its search results.
Or am I getting you all wrong? What am I not understanding about what you are saying?
So this week I bought a tractor after learning a LOT about which tractor was right for my needs from a LOT of other people that have experience I didn’t have. Then, I learned how to do some things I plan to do with this tractor.
Tractors and yard work aren’t exactly high-tech. Yet I found vibrant online communities and a lot of shared experience and knowledge. The web didn’t make me smart or stupid but it sure made it easier and faster to find PEOPLE that could make me smarter. The web and all the protocols, applications, services and even search engines on the internet are simply tools.
When I was young I didn’t get stupid when I got access to things like the public library where I could find and consume more information, faster and less expensively than on my own. The internet is no different.
I think his article is interesting and does not deserve so dismissive a response, sorry. How we learn does effect how we think. Search is changing our way of acquiring information, its value and meaning. It is something new for kids to do their homework at the feet of Google instead of pulling down a heavy volume of Britannica. The precision of text search is like a Cruise missile, obliterating the context one often would get in doing research. On the one hand. Google and other tools also open us to new horizons and information we could otherwise not have or bother finding. All this is changing us, I am just not sure how exactly and whether it is good or bad. But I would not dismiss Carr so quickly.
I’ve been posting a lot lately. Hope that’s not too much of a problem. I’ve yet one more comment:
“Yet, for all that’s been written about the Net, there’s been little consideration of how, exactly, it’s reprogramming us. The Net’s intellectual ethic remains obscure.”
While I agree with you that Carr should not immediately conclude that this makes us stupid, it is worth pointing out that there has been research into what effect some of the cog-like machinations on the web have on us.
In particular, you’ll notice that Google has made a conscious decision, on its home search page, to present the user with a single line text input area, instead of a multi-line text input box.
This is a subtle, but powerful, behavior-shaping interface. Researchers have studied this, and found that users tend to type shorter queries when give a single line, and longer queries when given a multi-line area:
http://www.ling.su.se/staff/franzen/irinterface.html
So already, Google is programming us, or conditioning us, to enter shorter queries. Rather than longer, more thoughtful ones in which we express our search intent with a little more clarity. And thought.
Does shorter necessarily imply dumber? I’m not saying that. But other studies have shown that longer queries have a positive correlation with both overall search task performance as well as end-user satisfaction with the information found.
But rather than encouraging users to enter longer queries, and thus achieve more satisfying results in the SERPs, Google sticks by the single-line input field, encouraging shorter queries, worse performance and, one may argue, more distraction.
And just as department stores that play slightly off-tempo music know, distraction leads to an increase in consumption. Or in this case, ad clicking.
I think Carr has got a point.
John,
Brains are plastic. They are modified by inputs. Too much shallow and broad information, without sufficient depth, will train brains to be shallow and broad, and lack depth of thought. The web hijacks our brain circuits that seek novelty.
I wrote a post a few years ago called “Is Concentration the New Competitive Advantage” and I still believe what I wrote. http://www.businesspundit.com/is-concentration-the-new-competitive-advantage/
Carr is exaggerating, but yes, there is a kernel of truth in what he says.
ummm . . . Nick’s blog reinforces his argument — he’s clearly been driven to distraction . . .
but seriously, as a recovering economist, I always look first for economic reasons for changes in people’s behaviors over time, rather than, for example, a technologically induced ADD; and I submit it’s harder to curl up for long stretches with War and Peace these days simply because the informational opportunity cost is higher (this is a long sentence — still with me?) — that is, I can now get all sorts of information any time from the Net that competes with any specific item of content I’m currently consuming. That opportunity cost of alternative information of a similar nature was not nearly as high in the past — Archie Bunker would be less likely to pull me away from my reading of War and Peace than my ability today via the Net to quickly delve into the details of Tolstoy’s life, and his circle of friends, and myriad other closely related or tangential tidbits.
Of course, Nick’s complaint about AI not being able in principle to handle or express ambiguity represents a vulgar Boolean view of the computationally possible — “maybe” is a definite possibility as an output from an algorithm.
Btw, agree with earlier comment that John’s 20% comment has nothing to do with Nick’s referenced point.
Yeah, like in the good old days when only Monks could read. Just another lazy peasant rant. (If we gave them plows they would only become lazier) And since I have spellcheck, my spelling has gotten worser.
So for example do math books in algebra still have page of logarithms in them and do we teach interpolation or deriving square roots by calculation ?
very insightful, thought provoking and interesting piece of writing.
it is the same busy mind as the yogis try to overcome, in the guise of information.
as vo nguyen giap said long ago about computers, they merely render your ignorance more efficient.
I think Carr is on to something. It’s very easy to be dismissive of Carr’s warnings in the same way we laughed at the admonitions against too much TV watching, but as I watch people around me increasingly drawn into mindless surfing, getting nothing done all day, and read a blog post by you, a professional writer, making basic grammatical errors you couldn’t have forced yourself to commit ten years ago, I have to wonder. Some of the other commentators here are correct, you shape your brain by Google result surfing the same way that you’d shape your body eating nothing but fast food. Believe it or not, there was a time when we celebrated the convenience and abundance of the McDonalds diet, and considered it the pinnacle of agricultural science. We know better now, don’t we? I suspect in 20 years we’re going to be bemoaning the effect of the Google intellectual diet.
John, I have been visiting your blog for the past couple of years. Your insights and opinions have been quite informative and inspiring.
I am part of a very small informatics company.
Our technologies are based on our experiences in the world of life sciences specifically chemistry informatics.
A couple of us decided to leverage some of
our technologies and experiment with
vertical search in the domain of music.
When you have a moment one day, check out http://music.zoowee.com. Your thoughts would be appreciated.
Thanks!
Good stuff John.. I agree with several of the points you raise, and I couldn’t agree more with the your last several sentences in the second to last paragraph.
No question in my mind that as we follow discussions from link to link and engage with people on any number of topics that we’re learning new information all the time. But stitching the conversation together takes some real effort, as we teach ourselves to filter through the noise that the proliferation of content brings upon us. We’re sharpening skills to cut through the clutter.
All that said, I agree with a lot of comments here as well: we shouldn’t be so quick to discard Nick Carr’s premise. The reality is that the Google Search model is affecting how we gather process information. I think in a lot of ways, it is similar to the big advances that he mentions that shaped new versions of the future. But it will take years for anyone to find evidence of just how it is affecting us.
information is not wisdom, why would you think otherwise?
If I know the name of a company that I am looking for, then I can enter that name into the Google lookup machine and it will spit out the domain name of the company (and it will work better if that company has been spammed with links first).
This can be useful if I don’t know the domain name of the company I have been looking for.
Google makes money by taking money from other companies to also show up when I am searching for the company / brand name which I actually want to find. This is a good idea, because that way, Google gets companies to comply with Google’s rules (kind of like a “performance” of pietà for the Roman Catholic church).
So what the other companies who are bidding on competing brand names are doing is to promote the brand name company to comply with Google’s rules (which Google constantly changes in order to be able to cash in more money). Dr. Seuss wrote a story about this type of money-making scheme called “The Sneetches” maybe about half a century ago — maybe people would do well to read that?
So Google is simply a simple algorithm plus a marketing plan. People who are smart will recognize that the advertisers who are wasting their money on Google (for example: by bidding on “BMW”) will achieve little more than promoting compliance with Google’s ever-changing rules.
Dumb people won’t recognize this.
Intelligent people will move on.
so, you think mediated reality consumption leads to wisdom? context? improved human values? an expanded heart? a more loving disposition?
i see no evidence of it in your article, especially the latter. i think you are more alone in your view than you might think.
the dominant intellectual ethic of the internet is that people are still consumers, meaning comes from brute force cataloging of letters and words, page-views is the golden land, and oh, yes, information is mostly free.
i think you were in a bad mood when you wrote this post.
For me this is timely and I agree 100% agree the web is not making us more stupid. I just had the thought that surfing the web and using Google and search can be seen as a sporting event for your mind- you start out by warming up with your favorite sites, start clicking links, doing some searches and suddenly you find yourself absorbed in an unexpected article and the in the zone like Kobe. Fire up some MP3’s for a little inspiration and then write a blog post about it all. Repeat and exercise the mind.
LOL — looks like I have managed to turn off ads competing with the BMW brand.
I guess you can send me a check c/o John Battelle…
;P nmw
I agree completely. The Web ecosystem is fueling, not draining, deep passion in the pursuit of knowledge. It spurs curiosity and inquiry. Yes, many links are dead ends. Yes, pages and blog posts can be the mental equivalent of candy. But just as often I find myself getting excited and inspired. A simple thread, a provocative idea, leads me to a deep dive. Today, thanks to blog posts and Amazon.com, I buy and read more books, not fewer. I also feel more connected to what I read. This is due, in no small part, to my newfound exposure to a community of like-minded readers. I now have easy access to an informed, and increasingly civil, debate on topics that interest me. I have tools to conduct my own process of personal discovery. Non-sequentially. Interactively. For me at least, compared to my more passive reading of magazines and books before the advent of the Web and blogs, that’s a huge improvement. I would not go back willingly.
The Web ecosystem is fueling, not draining, deep passion in the pursuit of knowledge. It spurs curiosity and inquiry.
So how often do you go past the first page of results on a “pursuit of knowledge” search that you do? Almost never, eh?
And how often do you enter more than 2.3 words, for one of your queries? Excepting those times when you are quoting a song title, to try and find the mp3, almost never, eh?
I’m just saying that the way we are being conditioned by the front-running search engines, e.g. Google, is to go very wide and shallow, rather than narrow and deep. Don’t find exactly what you want in the top 10? Try another 2-word query. Then another one. Then maybe a 3-word query. Then back to another 2-word query.
Heaven forbid that you actually try a 7-word query. Or that you look at the 5th or 6th page of results from a single query, in pursuit of a deeper understanding of a topic.
To me, that’s the issue here. Not whether the web helps us pursue knowledge. But whether the interfaces and algorithms offered by Google, the very structure of how they choose to elicit information from, and present information back to, the user are encouraging shallow exploration of the web knowledge, rather than deep, thoughtful exploration.
All the evidence I’ve seen points to the shallow.
So Google’s brand of search is geared to simultaneously program people on how to make short inquiries, while at the same time programning its users for diminished expectations. They also make tons of money by being the easiest company to do business with. And they are doing this during a time where the cost of information is trending downward towards zero, while Internet usage is trending upward with as many as 2 billion or so new minds coming online. It’s working for them – a 61% marketshare proves it.
There is something about that Google formula that is just so American – to me its like the invention of the assembly line by Henry Ford. The latter provided affordable transportation to the masses and more freedom to roam and explore America and fostered entirely new industries along the highways. The former provides almost free access to information – the freedom to select and access from the entire world’s information, fostering new industries along the pathways to that information.
Does Google have the vision to learn anything from Ford’s many mistakes?
Nick Carr’s viewpoint is wrong, of course, but not surprising. Print journalists have an economic interest at stake…the Internet is turning their business model upside down, and they are fighting back as hard as they can to protect their cushy jobs and lifestyles. I see the same behavior among journalists at the Washington Post here in DC. Classic Luddite behavior.
There is something about that Google formula that is just so American – to me its like the invention of the assembly line by Henry Ford.
Wow, I’ll have to admit, pointing out the sheer American-ness of Google is a really interesting exercise.
The more I think about it, the more I can see comparisons between Google attitudes and American attitudes. Only I don’t see Ford as the appropriate comparison. I see McDonald’s as the better analogy.
One of McDonald’s main drivers is speed. Sure, McDonald’s has to deliver a basic level of quality. The food has to be edible. But beyond that, quick ordering time (“Fun Meal #3, please!”) and quick delivery time (Americans did invent “fast food”, after all) are the quintessential McDonald’s experiences.
Same as Google!
Sergey Brin and Marissa Mayer have said, again and again over the years, how important a quick answer (quick delivery time) to a user query (a short one, because Google’s input box discourages longer ones, as discussed above — Meal #3, please) is among the most important features/factors driving Google.
As a result, the results served up by Google tend to be pretty shallow. The first page looks nice. But even with 1.2 million matches to your query, go to 3rd page of results and quality degrades fairly rapidly. That seems rather like a McDonald’s hamburger. That first bite tastes great, but after that it is a lot of empty calories.
“[Google] provides almost free access to information – the freedom to select and access from the entire world’s information, fostering new industries along the pathways to that information.”
Just as McDonald’s fostered (catalyzed) whole new industries, such as the potato industry in Idaho. And made a lot of people rich. Yay.
But at the end of the day, all those fries fill us with loads of transfats. They seem nutritious, and just as information obtained, two-word queries at a time, seems informative. But look at what eating McDonald’s for 30 days did to Morgan Spurlock. What does a year-long diet of nothing but Google-mediated, two-word shallow query information streams do to someone?
I agree. Google is very American. Like fast food is American. Cheap to consume, plentiful to its customers, but utilizing lots of resources to produce (server farm == livestock farm?) With hidden costs that can still be felt years down the road.
Google is to information retrieval as McDonald’s is to food. I like that analogy.
When is the Alice Waters of the search world going to appear?
Here I am!
😀 nmw
Gains in human intelligence are directly proportional to the rate at which information is distributed. With the birth of mail post, through the evolution of the wire to phone, radio, then television — come great gains of innovation and raw evolution in the form of advancement of our minds and our environments.
Only a pompous dolt looking for a soapbox would even dare to be so bold.
cidden cok güzel olmuş harika yazılar cıkıyor burdan.
Clearly, free access to the depth and breadth of the Web is a very good thing. But, without proper context and qualification, data is just data, and that’s where the issue lies, imho.
For those who buy into the keyword-box-as-Delphic-Oracle metaphor, use their 2.3 word queries, click on the first 2.6 blue links from Google (or Ask, or Msft, or Yhoo) and think they see “the truth”, well, maybe search isn’t going to help them them get into Mensa anytime soon.
But, for those who have some understanding of the tool they use – who understand that those 10 links Google gives you are but one possible solution to a hugely complex multivariate optimization problem, not THE solution, search can be a hugely powerful window into the Web.
Of course, given human nature and our strong bias towards satificing behavior, I’ve got to think the vast majority of search users fall into the first camp, either because they don’t know any better or because they are too lazy to look further.
Hmmm … looks like I’m right back at the Google as McDonalds analogy that JG riffed on so eloquently … what he said!
And how often do you enter more than 2.3 words, for one of your queries? Excepting those times when you are quoting a song title, to try and find the mp3, almost never, eh?
And it fits into their business model as well. It makes sense to “program” web users to exhibit short attention spans and shallow behavior when searching, increasing their search frequency (when you cant find what you want in the first couple of pages, you finetune your query and do another search, which in turn results in two things – (1) much more information and profiling of user searche behavior and (2) much more ad space inventory, views and higher probability of clicks. This being a network economy this translates into more revenue.
The internet is a world full of promise, blogs, and scams. People get caught up in everything they google, ask or yahoo. Men and women rely on the internet to answer all questions and direct them in the right direction. One half of me believes people depend to much on the internet world. Where would we be without it though?
This is not totally unrelated to a post I wrote recently at http://corp.viewzi.com/index.php/v2/entry/the_search_for_unsearchable_searches/ on the opportunity that can be created by the programming of the search engine user BY the search engine.
Don’t you think that this post, which completely dismisses what would otherwise be a salvageable point and which is written by someone with a very clear bias on the subject, somewhat proves the point that the internet is not the height of intellectual pursuit? When the experts resort to sarcastic remarks and sophist retorts with very little deep thought put to the matter, then you have something dumber than books typically provide. The higher barrier to participate in the academic debate that exists with books does do a decent job of filtering out that long tail of BS you get with blogs.
I have to agree with Nick: he is a contrarian, conservative link-baiter — but this makes understanding Internet much more fun.
Any time he makes a point like this (quite often) I can’t help thinking of “This will kill that”, a chapter in Notre-Dame-de-Paris (Yes, the same book that Disney crushed into grotesque.) where Frollo explains how print will over-come architecture as a mean of expression, reshuffling powers (“this” being the newly printed book that Frollo holds, and “that” the cathedral).
Internet makes any day of work feel like a cocktail party, or a conference: lots of noise, or what sounds like it, but is actually people meeting other people, trying to figure this out.
JG mentioned:
> But other studies have shown that longer queries have a positive correlation with both overall search task performance
That actually has a peak at three keywords; Google studied that extensively.
For those of you Mac users who feel the need to take the plug away, in a symbolic way, try that:
http://www.ibiblio.org/fred/freedom/
AJ: I read your link. Really, really good points. I completely agree:
“The more you use the search engine, the more you understand what kind of questions can be asked and how they ought to be formulated to get optimal results. At some point, you stop attempting questions that you have seen fail in the past–whether the failure was a result of a lack of information, or the question just wasn’t suited for search engines at the time. As users are trained on limitations, the demand for answers to the questions where the search engine is weak diminishes, and so there is a less apparent urgency for the search engine to improve where it is weakest. This transactional relationship could create blind spots for major search engines”
You come at it from a developmental psychology standpoint, looking at the user. I come at it from a machine learning / probability standpoint, looking at Google.
Google has this strong cultural leaning toward only reacting to the user, to developing its algorithms and interfaces based only on what it sees the user do. It’s part of their whole cult of engineering, where everything is measured, and engineers prove their ideas not through rational, intuitive, or insightful arguments, but through cold, hard measurements of the numbers. At least that’s what they say all the time, through their PR outlets
So the search engine itself gets re-engineered to fit exactly and only what the user is doing. And if the user is no longer attempting certain things, as you say above, then the search engine will never be able to measure those things. And by not being able to measure, no solutions will ever be designed and implemented. And with no solutions available to the user, the user is going to be even less likely to try those things. Vicious downward spiral.
It’s what we in machine learning call a “local maxima”. If your objective function is only what the user *is* doing, rather than what the user *could* do, if they were trained better, then you are going to engineer your system to iteratively hill-climb your way to a sub-optimal solution. A local maximum.
But other studies have shown that longer queries have a positive correlation with both overall search task performance
That actually has a peak at three keywords; Google studied that extensively.
Bertil, could you provide citations? Because the papers I’ve read show that results continue to improve even with 10 and 20 keywords.
I presume, JG, that Bertil meant the default “AND” search (or perhaps even a phrase search)?
As we both know, there are many “armchair professors” out there with little knowledge of advanced search methodologies.
That said, I too would appreciate at least 1 citation (and yes: a simple link will suffice).
Strange how the Internet has changed: in the early days, the academic community would never have even thought of setting up any hypothesis with supporting documentation (I write more about this in http://gaggle.info/miscellaneous/articles/wisdom-of-the-language )
jg, no link to you, hope you read this here
you are making a good point, and it is like science, which can only ask those questions for which it can find proof, and thereby ignores huge swathes of reality….
google mind is going to prove to be very limited
@gregory: I know and accept that, in science, there are certain questions that we will never be able to find proof (or, technically, “find disproof”), simply due to the nature and structure of the problem. Goedel’s incompleteness theorem lets us know that.
But I’m not talking about those sorts of problems. I’m talking about the fact that there are things for which Google has the ability to find proof.. but doesn’t want to because.. well, who knows why for sure. Maybe it’ll disrupt their business model?
@nmw: I don’t know if your call for a link was to Bertil, or to me. In case it was the latter, here you go:
http://sils.unc.edu/~dianek/belkin-sigir03.pdf
Citing the abstract:
Query length in best-match information retrieval (IR) systems is
well known to be positively related to effectiveness in the IR
task, when measured in experimental, non-interactive
environments. However, in operational, interactive IR systems,
query length is quite typically very short, on the order of two to
three words. We report on a study which tested the effectiveness
of a particular query elicitation technique in increasing initial
searcher query length, and which tested the effectiveness of
queries elicited using this technique, and the relationship in
general between query length and search effectiveness in
interactive IR. Results show that the specific technique results in
longer queries than a standard query elicitation technique, that
this technique is indeed usable, that the technique results in
increased user satisfaction with the search, and that query length
is positively correlated with user satisfaction with the search.
And the final paragraph from the same aforementioned paper:
Thus, we conclude that our quite simple interface-based query
elicitation technique results in significantly longer, and more
useful searcher queries in a Web searching task than typical
query elicitation, for a best-match information retrieval system.
Furthermore, we conclude that longer searcher queries result in
increased search effectiveness in general, indicating that more
words from the searcher describing the person’s information
problem results in better interactive IR performance. Taken
together, our results mean that getting longer queries from
searchers in a best-match Web searching environment is not
only possible, but desirable and useful.
This paper was published 5 years ago. 5 years. And what has Google done, in the meantime, to change its interfaces and algorithms to elicit longer queries, to improve user satisfaction and performance, so that users more often find what they need in the SERPs, rather than resort to clicking an ad? As far as I can tell, nothing. The Google interface looks exactly like it did in 1998.
Am I really expected to believe that the first interface that Google ever tried, over a decade ago, is really the best one that will ever be? That Google has tried* implementing query interfaces that elicit longer queries, and that every single one has failed, lending evermore credence to the belief that the 1998 interface is already perfect?
If we believe that, then Nick Carr is correct. Google has made us stupid.
—-
*(And if so, why have I never, in 10 years of using Google, seen any of these other interfaces? They were never tested on *me*!)
Thanks, JG!
Yes, I totally agree.
Have you tried http://quintura.com ?
I think that site is alot of fun…. (*)
😀 nmw
(*) but actually I easily tire of keyword clustering :S (but still: the way they have done it is — I think — like *way cool*… how long before somebody buys them? or makes an agreement to “protect” their property? ;P)
I agree with JG, “I think there is some merit to the point that Google makes us dumber, because it conditions us to believe that its top 10 results are (1) the best answers, and (2) the only answers. Even if it says that there are 1.2 million hits, Google gives us no informational, sensemaking instruments to help us deal with those 1.2 million hits, and thereby encourages us to lazily not bother. Who can possibly scroll through 1.2 million results, 10 links at a time? And even if you do have the patience to manually slog through, say 500 of those 1.2 million results, how does Google really help you organize and categorize and sort through what you’re finding, as you find it?”
Hmm… I read the story from The Atlantic, and I couldn’t find Nick Carr stating that “Google makes us stupid”.
Was it there somewhere between the lines?
Anyway, I’d see that the web is clearly one of the most human technological inventions. One example: we don’t have to learn arcane commands and syntaxes to use it. There’s a big difference between building an SQL query and using web search.
Furthermore, what’s the difference between reading a book (i.e. going through a set of pages), and browsing the web (likewise: going through a set of pages)?
Oops, can I fix the URL? Thanks.