How Search Engines Have Changed the Way We Learn

Most people don’t stop and think about how to find information online or how to navigate to a specific website anymore. They open up their web browser, type any related terms into the text bar, and hit “Enter”. Though this seems like it has been the natural order of searching for information for some time now, the advent of the search engine is relatively new. Before the search engine was popularized, exact website names had to be known in order to be directed to them. This shift towards mass search engine use has created a cruder form of information mining, where the user instead searches for vague ideas according to keywords instead of using exact terms, or being directed precisely to the page they want to view through knowing the website URL. This shift has greatly changed the way that society searches for and researches information. The purpose of this research paper is to explore how search engines have changed learning and cognitive processing of users.
To understand how search engines have affected society, it’s important to understand the history of search engines. The first technological ancestor of the search engine actually starts long before computers, with the library catalog. Library catalogs are not a modern invention by any means, as they originated in ancient Mesopotamia in the 7th century B.C. (Mason, 2017). They have evolved from wall plaques to scrolls, to books, to card catalogs, and now computer databases. However, modern cataloging techniques were developed by librarian of the Smithsonian, and later of the Boston Public Library, Charles Jewett. Jewett was perhaps one of the first proponents for a national library, and he designed a system of searching based on authors and titles. However, it was Jewett’s coworker and understudy Charles Cutter who was concerned with organization for the customer who was searching for a subject without knowledge of the author or the title. His motto was:
The convenience of the public is always to be set before the ease of the cataloguer... [saying that] a plain rule is not only easy for us to carry out, but easy for the public to understand and work by (Cutter, 1904).
Cutter’s advanced system of classification marked a shift in the manner that the public could find content. No longer did those searching for a book on a particular subject have to know the title and author, they could now search by subject according to Cutter’s system. Imagine if, in today’s society, a person were trying to find a scholarly source online concerning a research topic, but were only able to search according to the title and authors that they knew? It is obvious that Cutter’s library classification system paved the way for modern search techniques. There have been many advancements and variations of the library catalog; nevertheless, Cutter’s classification system of subject headings is still used by the Library of Congress today.
However, no change was quite so revolutionary to the development of search engines and database cataloging as the introduction of a text retrieving program known as SMART. SMART was developed in the 1960s by the father of modern search technology, Gerard Salton, and it was used as the foundation for modern search engines (Consroe, 1996). Even before the internet, there were private networks, databases and archives (mostly used by the government and large libraries) that had digitalized documents that needed to be stored and easily searched. Due to this development in computer technology and new media to categorize, even Cutter’s classification system was deemed outdated. One of those to raise awareness of the effect that digital catalogs were having on the traditional library classification was Patrick Wilson. In 1983, regarding the rise of computer technology, Wilson says in his article, The Catalog as Access Mechanism:
Computer catalogues have made main entries [author and title] obsolete; altered how they make entries; and transformed catalogues into search stations that can use various prompts to retrieve information (Wilson 1983, 261-262; Mason, 2017)
Salton’s SMART technology paved the way for what many people consider the first search engine: Archie. Archie was an archives program applied to the Advanced Research Projects Agency Network or “ARPANET”. ARPANET was created in 1969, and was the network which eventually led to the internet. On top of this network structure with the introduction of Archie in 1990, Tim Bernes-Lee designed the World Wide Web in 1991. Bernes-Lee’s network featured a page that contained an organized virtual library with a catalog page logging and linking to the most popular, useful websites of the time (Wall 2006). It is evident that without traditional catalogs and archival methods, there would be no foundation upon which to base the modern search engine.
Some of the benefits of search engines are evident. Users are now connected to many sources that may have been inaccessible without the knowledge of the exact website name. Because search engines are an inexact, crude method, there are many options for searchers—this is helpful to those who may not even know precisely for what they are looking. Continuing the metaphor of information mining, the “miner”, or the searcher, encounters both material that is useful, and useless. It is the searcher’s job to distinguish the gold from the dross and to extract it. The search engine casts a broad net, and allows the “fisher” the power to choose the material that is “meatier”—or of more valuable substance. The search engine gives the information seeker not only the power of choice, but the power to determine what information is of value. This resource’s value is assigned not only according to the individual user, but also to the community that uses search engines, as the algorithm places the resource higher on the list due to increased selection. Obviously, the curation of quick, applicable information is the purpose and benefit of a search engine. In relation to Gestalt psychology, many would consider these benefits the “figure” in the figure-ground perception theory (Moore, 1993).
But what are some of the unseen effects of this technology? The ground of search engine technology is far less obvious. However, limitations in the realm of text and data mining are the same limitations that search engines have brought to the academic world. To understand what is meant by this, it is important to know what is meant by data mining. Data mining as defined by the dictionary is “the process of collecting, searching through, and analyzing a large amount of data in a database, as to discover patterns or relationships” (Data Mining, 2017) According to Dr. Jane Yung-jen Hsu, chairwoman of Computer Science and Information Engineering at National Taiwan University, there are three things that must be considered with any informational documents that are mined. These three requirements are:
1. Content: the actual data in a document
2. Format: the visual presentation of the document
3. Structure: the logical elements and their relationships (Hsu, 1997)
Search engines in our modern society have become part of the first steps of data mining. One of the benefits of search engines has become one of the negative aspects as well: search engines are only able to search by content via keywords. This is often why this “mining” method is crude—search engines completely ignore both the format and the structure of the document. Format and structure are both vitally important as they often indicate the relationships of different terms and ideas. Though some results may be filtered according to the frequency of its use, search engines are only the first crude steps in gathering useful information. Dr. Yung-jen Hsu likened this to an owner talking to a pet dog (Hsu, 1997). For example, in the sentence: “Fido, I told you not to dig in the trash!” The dog doesn’t understand the full idea that the owner is trying to convey. The dog may understand that it is being referred to, or addressed, after the owner yells “Fido,” but the structure is lost. Search engines have this same limitation by searching with keywords. Even if more keywords are added to create a relationship between terms, there are limitations with this method. Suppose Fido’s owner were to yell “Fido, I told you not to dig in the trash,” and then follow up with the statement, “you won’t get a treat tonight!” Assuming Fido has been conditioned to do so, he now hears the word “treat” right after his name. It is easy to understand in this context that Fido might have the wrong conclusion of what message the owner is trying to convey. So it is with materials found via search engines. Keywords often do not provide enough context of the relationship between the two terms to be useful. Therefore, the user of the search engine often has to deal with sub-par sources when referencing for scholarly works.
To combat this problem, search engines use text clustering. Clustering as defined and described by Charu Aggarwal is:
...finding groups of similar objects in the data. The similarity between the objects is measured with the use of a similarity function. The problem of clustering can be very useful in the text domain, where the objects to be clusters can be of different granularities such as documents, paragraphs, sentences or terms. Clustering is especially useful for organizing documents to improve retrieval and support browsing. (Aggarwal, 2012)
Because clustering finds groups of similar objects in the data it can simulate structure of a source or document. Still, informational clustering cannot determine relationships as well as cognitive intelligence decoding raw data. The format of the source material also indicates relationship, and clustering does not factor in this parameter. Informational clustering in search engines is fundamental. Clustering only recognizes relationships that have already been established, but it takes ground level research in order to establish new relationships. Research by methods found by pure “clustering” methods results in inadequate and second-rate material. Sub-par material often tries to “force” a relationship between two variables because the data may be related through keywords, but seldom does it apply directly to the problems that the researcher is addressing.
Research via clustering algorithms found in search engines proves an important negative aspect of society: real research methods have largely been forgotten. Additionally, because search engines have made searching by keywords extremely popular, this lackadaisical search method has carried over into the area of research, leading to substandard research methods. Search engines do facilitate research, but too often in modern academic society, that is where “research” ends. However, search engines’ failure to result in qualitative research is not the only adverse effect that this medium has had on society. Society’s consciousness itself has been affected, leading to a phenomenon known as the “Google Effect”. According to International Business Times:
With so much information available, there is less need to remember everything, especially with tools like Google allowing us find what we need quickly. The result–the Internet becomes an external memory for humans. (Davis, 2012)
The original conceivers of this notion, researchers Betsy Sparrow, Jenny Liu, and Daniel Wegner, made an even more alarming statement:
We are becoming symbiotic with our computer tools, growing into interconnected systems that remember less by knowing information than by knowing where the information can be found. (Sparrow, 2011)
In the same way that clustering can’t create new relationships, neither can humans because of their reliance on a medium that deals with information that is already established. What this research shows is that humans are becoming passive agents of information—a conduit in an information system—rather than active original thinkers. There is no “grunt work” involved in the research process, because society is drowning in information, and has little reference point as to what is important information and what is not. Though he doesn’t reference research through search engines, Douglas Rushkoff addresses this oversaturation of information in his book: The Present Shock. Cluster thinking creates a society where everything is connected, and those caught in middle can’t differentiate raw source material from the effects or “feedback” that the information creates. Rushkoff writes:
When feedback comes instantaneously and from all sides at once, it’s hard to discern what [the reaction is]…so we create context via links (Rushkoff, 2014)
These “links” are the dangerous part of the Google Effect. We are no longer self-sustaining in our cognitive abilities or understanding. Clusters, both in search engines and cognition, are dangerous, because they create algorithms to predict and infer information without the ground level research to assure actual relationships exist between factors. Clusters are shortcuts for learning.
In the original research of the Google Effect, participants could remember “where” information was found as opposed to the specifics of the content. This is a form of transactive memory, or a “symbiotic relationship” with the search engine (Sparrow, 2011). The society, in adaption to the digital frontier, has traded retained information for neural shortcuts to the source material. In the study “Decisions and the Evolution of Memory,” researcher Stanley Klein states:
Memory evolved to supply useful, timely information to the organism’s decision-making systems. Therefore, decision rules, multiple memory systems, and the search engines that link them should have coevolved to mesh in a coadapted, functionally interlocking way. (Klein, 2002)
Society has evolved to replace stored information with neural pathways to external information—almost like a computer shortcut to an external drive. Because there is mental acknowledgement of information that exists on a subject elsewhere that is accessible, society doesn’t have a reason to hold long term memory of information. This phenomenon has negated memory itself. According to psychologist and neuroscientist Endel Tulving:
Memory in biological systems always entails learning (the acquisition of information) and ... learning implies retention (memory) of such information. (Tulving, 2004)
What Tulving makes clear in this statement is that rather than aiding in the learning process, search engines are actually hindering learning. Learning, according to the dictionary, is defined as “the act or process of acquiring knowledge…” (Learning, 2017). If learning is a process, then search engines subtract from the process of linear thinking. Because the process is negated, “acquiring knowledge” is then deemed impossible. If memory requires retention of information, then search engines are removing our ability to remember usable facts and knowledge. Because search engines have replaced critical thinking and deep exploration of the relationships between ideas, information and variables, society and researchers have become lethargic, to a point where, without search engines, research doesn’t exist. Even within search engines, research has become lazy. Joel Achenbach of the Washington Post responds to this lackadaisical research shift in an interview saying: “hardly anyone really knows how to do research online any more than they know how to do it in libraries” (Achenbach, 2004a).
Society has lowered its standards—or perhaps mistaken the definition—for research. This is due not directly because of search engines, but the shift in linear thinking to cluster or network thinking that has been caused by search engines. Search engines have created a society that attributes relations based on keywords and vague ideas rather than logically thought out progression of thought. This is not to say that search engines do not have benefits. Obviously, for quick information bits, search engines are extremely helpful. As Paul Saffo, research director for the Institute for the Future states in an interview with Joel Achenbach: “You’re getting the advantage of the group mind.” However, the “group mind” is not always the best approach. Saffo continues:
"The more people get on the Web, the more the Web becomes the vaster wasteland that is the successor to the vast wasteland of television. I don't care what the majority of people are looking at, because the majority of people are really boring," (Achenbach, 2004b)
Saffo puts into colloquial terms what modern researchers overlook—research that is already pre-packaged is not cognitively stimulating. Searching via the hive mind does not assist in original thought, and society may be losing its critical thinking ability because the knowledge to which they are connected isn’t their own.
This research does not necessarily indicate that search engines are negative for society. As all technology, it has benefits and disadvantages. It also has foreseen effects and unforeseen effects. Even as early as ancient Egypt, technology that makes information accessible through media other than the person was condemned. In the Egyptian legend of Thoth and Thamus (as told by Plato and recorded by Socrates), King Thamus disapproves of the technology of writing presented to him by the Egyptian god Thoth, saying:
This invention will produce forgetfulness in the minds of learners, with the neglect of their memory, because their trust in writing comes from strange external marks and not their own internal recall. The drug you’ve discovered isn’t a memory-enhancer, but a mentioning-enhancer, and you’re offering your students pretend wisdom, not truth. (Moore, 2012)
How pertinent Plato’s understanding is to the modern discussion of the effects of search engines! As with the prerequisite technology of writing, search engines obviously fulfill a great purpose, and have wonderful benefits to society. Perhaps the advent of a new technology will always come with opposition; nevertheless, it is important not to dismiss these warnings. Search engines are a tool for society to use, but to use it incorrectly can and will result in society that depends on “pretend wisdom”. It is the job of the researcher to do grunt work, and to ensure that the circulated “truths” and relationships as indicated by search engines have legitimacy, and are not solely the result of groupthink or a simulated understanding based in clusters. Still, the very name of the “internet” or “World Wide Web” indicates a network of understanding. Learners must find the checks and balances in their own lives that yield to development of themselves—not the indoctrination or loss of themselves based on algorithms and popular understanding found in search engines.
 
Sources:
Achenbach, Joel. “Google's Effect on Everyday Life.” The Washington Post, WP Company, 20 Feb. 2004a, www.washingtonpost.com/wp-dyn/articles/A55288-2004Feb19.html.
Achenbach, Joel. “Search For Tomorrow.” The Washington Post, WP Company, 15 Feb. 2004b, www.washingtonpost.com/wp-dyn/articles/A42885-2004Feb14_3.html.
Aggarwal, Charu C. Mining Text Data. Kluwer Academic Publishers, 2012.
Consroe, Karla. “IN MEMORIAM.” In Memoriam, Cornell University, 1996, www.cs.cornell.edu/Info/Department/Annual96/Beginning/salton.html.
Cutter, C.A. 1904. Rules for a Dictionary Catalog: Selections, pp. 62-71 in Foundations of Cataloguing, ed. by M. Carpenter and E. Svenonius. Littleton: Libraries Unlimited, 1985.
“Data Mining.” Dictionary.com, Random House Inc., 2017, www.dictionary.com/browse/data-mining.
Davis, Johnathan. “Google Effect: Changes to Our Brains.” International Business Times, 17 Dec. 2012, www.ibtimes.com/google-effect-changes-our-brains-299451.
Hsu, Jane Yung-jen, and Wen-tau Yih. “Template-Based Information Mining from HTML Documents.” Association for the Advancement of Artificial Intelligence, National Taiwan University, 10 June 1997, archive.agent.csie.ntu.edu.tw/~yjhsu/courses/u1760/Online/2002/aaai97.pdf.
Klein, Stanley B., et al. “Decisions and the Evolution of Memory: Multiple Systems, Multiple Functions.” Psychological Review, vol. 109, no. 2, 2002, pp. 306–329., doi:10.1037//0033-295x.109.2.306.
“Learning.” Dictionary.com, Random House Inc., 2017, www.dictionary.com/browse/learning.
Mason, Moya K. “Historical Development of Library Catalogues: Their Purpose and Organization.” Historical Development of Library Catalogues: Their Purpose and Organization, Moya K. Mason, 2017, www.moyak.com/papers/history-library-catalogues.html.
Moore, Christopher. “The Myth of Theuth in the Phaedrus.” Plato and Myth, vol. 337, 2012, pp. 279–303. Mnemosyne, Supplements, doi:10.1163/9789004224360_016.
Moore, Patrick, and Chad Fitz. “Gestalt Theory and Instructional Design.” Journal of Technical Writing and Communication, vol. 23, no. 2, 1993, pp. 137–157., doi:10.2190/g748-by68-l83t-x02j.
Rushkoff, Douglas. Present Shock: When Everything Happens Now. Current, 2014.
Sparrow, B., et al. “Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips.” Science, vol. 333, no. 6043, 2011, pp. 776–778., doi:10.1126/science.1207745.
Tulving, E. Memory: Introduction. Gazzaniga, Michael S. The Cognitive Neurosciences. Cambridge, Cambridge: MIT Press, 2004.
Wall, Aaron. “History of Search Engines: From 1945 to Google Today.” Search Engine History, 2006, www.searchenginehistory.com
Wilson, P. 1983. The Catalog as Access Mechanism: Background and Concepts, pp. 253-268 in Foundations of Cataloguing, ed. by M. Carpenter and E. Svenonius. Littleton: Libraries Unlimited, 1985.