The rise of the Hyperlink map
Using the the webcrawling tool Hyphe we have gathered data around the controversy topic IQ and what it depends on. In this way we traveled along the interconnecting hyperlinks of blogs and websites relating to or discussing IQ dependency. Doing this to find out which people and public figures of relevance communicates and discusses this controversy.
Crawling through the web we gathered an extensive network of several thousand unique websites only stopping the accumulation, when the websites we found no longer mentioned IQ dependencies causes or intelligens.
Using the crawling tool we sorted out anything with but one reference from other sites in the network, as it by virtue of its unpopularity was considered of little relevance to the other actors in the network.
This also removed the majority of the actors and made the next step in the process more manageable.
Finally in the data gathering and construction phase we started adding tags to the websites of the network individually, going from most cited to less. Doing this we gained a descriptions of all the major and some of the minor, actors of the network. This made it possible to find patterns and eccentric or irregular nodes that we can the analyze and discuss, as will be done underneath. Though not all the tags will be used or explained we will leave the network data available to anyone who wants to dig into it at the bottom of the page. In this way everyone can investigate their own thought and concepts to discover new findings.
Crawling through the web we gathered an extensive network of several thousand unique websites only stopping the accumulation, when the websites we found no longer mentioned IQ dependencies causes or intelligens.
Using the crawling tool we sorted out anything with but one reference from other sites in the network, as it by virtue of its unpopularity was considered of little relevance to the other actors in the network.
This also removed the majority of the actors and made the next step in the process more manageable.
Finally in the data gathering and construction phase we started adding tags to the websites of the network individually, going from most cited to less. Doing this we gained a descriptions of all the major and some of the minor, actors of the network. This made it possible to find patterns and eccentric or irregular nodes that we can the analyze and discuss, as will be done underneath. Though not all the tags will be used or explained we will leave the network data available to anyone who wants to dig into it at the bottom of the page. In this way everyone can investigate their own thought and concepts to discover new findings.
Enviroment vs Genetics and the use of statistical data
The following is an analysis of a network of sites from the world wide web concerning the controversy of IQ dependency. The network is illustrated bellow in map no. 1. In general the network looks like a hairball of interrelated links, due to the many connected links in a small space. The illustration however is still a good indication of the characterization of the the debate, surrounding IQ dependency and the following use of IQ. The debate is simply a heated and messy topic!
But why is this the case?
But why is this the case?
The Authorities When a network map of an online debate is created it can be expected that there are lots of referencing going into certain types of websites - such as government sites and published documents in news articles and scientific journals. These references are made from sites supporting and representing different publics and opinions on in and surrounding the topic. In map no. 1 we see a gathering of the largest authorities who interestingly are all placed in the same place, on the right of the hyperlink network.
The references is often used to legitimize an opnion. In this sense when an scientific journal or an authority is backing up an actor's positions, it appears more valid. This seems to be the case in our controversy, where blogs link to many authorities. What is more particular about this controversy is the high degree of interconnectedness in general. Normally in a controversy it is possible to filter out much of the hairball structure by removing the authorities from the corpus. By this we mean to remove the nodes that are extensively linked and are used as authoritative at the web. Doing this would normally result in spreading out the network into more issue specific component-clusters or groups. Following the advice of the research engineer Mathieu Jacomy, the top layer of the network was filtered out. In our case the National Institute of Health (NIH) was far the most well connected node along with the scientific journal Nature. These together with the other websites used as authorities was filtered out. Removing these authorities in the network however, does next to nothing to stifle the hairballing effect. This is illustrated at the right at map no. 2. |
The effect of blogs and interconnectedness
Another typical cause for hairballing is the effect of big hub sites such as blogs. These serve a hub effekt by linking to many different nodes in the network and thereby creating a hairball effect in the network. Following Jacomy’s general advice these hubs could be removed to make the network more specific. Again this have no effect on this network’s hairballing however and what is worse it removes most of the network, since most of the discussion of our controversy happens in the blogosphere. Blogs are so to speak the main actors in this controversy publicly.
A reason for this high interconnectedness between the adversary blogs of the controversy is for example "“Vdare” and “Pumpkinperson”" or "“Abagond” and “Notpoliticallycorrect”". Abagond is in opposition to Notpoliticallycorrect, but does still link quite a bit to it, for example to argue against their side of the controversy. You could say that the blogs need each other to position themselves and thereby create a distance to the adversaries and the public they represent.
A reason for this high interconnectedness between the adversary blogs of the controversy is for example "“Vdare” and “Pumpkinperson”" or "“Abagond” and “Notpoliticallycorrect”". Abagond is in opposition to Notpoliticallycorrect, but does still link quite a bit to it, for example to argue against their side of the controversy. You could say that the blogs need each other to position themselves and thereby create a distance to the adversaries and the public they represent.
So what does the actors think IQ depends on?
One of interesting prevalent tags in our network is that of genetic vs environment influence. Genetic is the most prevalent one including around 75% of all the tagged nodes in this map tagged by either environment or genetic influence.
This makes Genetic influence on IQ the most popular opinion in our network. As a comparison 45% of the tagged websites mentions environmental influence as an important descriptive factor for IQ. The reason for this summing over 100% is that 22% of our nodes argues for both genetical and environmental influence on IQ, and some does neither but still shows in the network due to being referenced by multiple actors. The hyperlink map is sorted by in-degree, meaning the bigger the size, the more hyperlinks goes into the node.
There is no distinct groups, which is a quite interesting point. One would expect that genetics and environment would be in different camps arguing between themselves, not wanting to give any web traffic to the opposing view. This is not the case, they link to each other and acknowledge each other.
This makes Genetic influence on IQ the most popular opinion in our network. As a comparison 45% of the tagged websites mentions environmental influence as an important descriptive factor for IQ. The reason for this summing over 100% is that 22% of our nodes argues for both genetical and environmental influence on IQ, and some does neither but still shows in the network due to being referenced by multiple actors. The hyperlink map is sorted by in-degree, meaning the bigger the size, the more hyperlinks goes into the node.
There is no distinct groups, which is a quite interesting point. One would expect that genetics and environment would be in different camps arguing between themselves, not wanting to give any web traffic to the opposing view. This is not the case, they link to each other and acknowledge each other.
Finding the active actors
Another way to investigate the network is as illustrated by map no. 3 . This map is made by sorting via out-degree, that is making the node size larger the more the site links out.
This is a way of finding hubs, but it also makes it possible to see which actors actively engage with others in the network, as in trying to write for a viewpoint against another site or attempting to mobilize a recognized source to support their own opinion. The now much smaller dots with many lines going into them, are in many cases the former authorities. These are government websites or journals containing scientific articles such as Sagepub or Nature, sites who do not themselves engage in the debate but are useful for those who does. Sagepub and Nature are coloured by the articles linked to in their database and the color is thus not an indicator of official position. The active actors in this illustration is mainly the blogs such as "Abagond", "Notpoliticallycorrect", and "Thealternativehypothesis". A presentation of these actors is find at the Lippmannian section. |
Statistical data, anti science and a diverse environment group
A further investigation of our map can be made using the tags genetics contra environmental in comparison with the tag statistic data. It was found that the blogs using and describing statistical data are all, without exception describing genetics as primary influence of IQ. Only one of the blogs describing statistic data, Jonahlehrer, mentions environment as a relevant factor for IQ as well. This indicates that statistical data seems to be quite relevant form of argumentation for those supporting the genetic explanation, and on the other side it seems to be quite irrelevant as evidence for those arguing for environmental effects.
Reverting to the blogs asserting “genetic influence” on IQ dependency, we find four of these sharing the tag “anti science“, which is a tag created if the blogs mentions the words “anti-science” about their debate opponents. In these cases it is used about the “environment” tagged blogs as standing in the way of science due to their lacking respect of statistical method. It also demonstrates a big issue that both groups raise. This is the politicization and moralisation of science. Both sites are against science being politically motivated, and use either the word “anti science” or “racist” about each other, blaming the other public for having a political agenda behind their arguments. It also illustrates the relevance of trust, by blaming the other side for having a political agenda, they automatically distrust their arguments and claims.
While a smaller part of the overall network consists of blogs describing environment as the most determining factor for IQ, these blogs seems to be more spread out in their positions as based on the tagging we conducted. For example we find blogs such as timewise and themonkeycage to be tagged with the tag “against IQ-testing” meaning that they do not consider IQ a relevant tool for measurement. Another example is that under environment different sites references different causes of environmental influence on IQ. Some say that education is the most important, others family upbringing, and third nutrition. We also find positions via tags such as “no race” meaning that they argue for that race is an artificial construction which therefore is irrelevant to anything. Especially race is a hot topic: on one side we have Timwise.org and Abagond.Wordpress who both criticize the very concept of race as a social construct and therefore irrelevant. These two argue for primarily environmental influence on IQ. On the other side sites such as Vdare and Notpoliticallycorrect argue for race having a huge if not all imposing influence on IQ discrepancies, and therefore consider it of the utmost importance.
Reverting to the blogs asserting “genetic influence” on IQ dependency, we find four of these sharing the tag “anti science“, which is a tag created if the blogs mentions the words “anti-science” about their debate opponents. In these cases it is used about the “environment” tagged blogs as standing in the way of science due to their lacking respect of statistical method. It also demonstrates a big issue that both groups raise. This is the politicization and moralisation of science. Both sites are against science being politically motivated, and use either the word “anti science” or “racist” about each other, blaming the other public for having a political agenda behind their arguments. It also illustrates the relevance of trust, by blaming the other side for having a political agenda, they automatically distrust their arguments and claims.
While a smaller part of the overall network consists of blogs describing environment as the most determining factor for IQ, these blogs seems to be more spread out in their positions as based on the tagging we conducted. For example we find blogs such as timewise and themonkeycage to be tagged with the tag “against IQ-testing” meaning that they do not consider IQ a relevant tool for measurement. Another example is that under environment different sites references different causes of environmental influence on IQ. Some say that education is the most important, others family upbringing, and third nutrition. We also find positions via tags such as “no race” meaning that they argue for that race is an artificial construction which therefore is irrelevant to anything. Especially race is a hot topic: on one side we have Timwise.org and Abagond.Wordpress who both criticize the very concept of race as a social construct and therefore irrelevant. These two argue for primarily environmental influence on IQ. On the other side sites such as Vdare and Notpoliticallycorrect argue for race having a huge if not all imposing influence on IQ discrepancies, and therefore consider it of the utmost importance.
dynamic maps
Underneath you find a dynamic version of the maps used in the analyses. This create a possibility for the reader to investigate and move around in the maps.
Please notice that all the maps are showed as an in-degree version.
This dynamic version is presented via the tool manylines.
Please notice that all the maps are showed as an in-degree version.
This dynamic version is presented via the tool manylines.