Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!
GIST: An Interdisciplinary Perspective
![Hillary G's picture Hillary G's picture](https://serendipstudio.org/exchange/files/pictures/picture-790.jpg)
Hillary G
May 7, 2011
How Gender, Information, Science, and Technology Interact:
An Interdisciplinary Perspective
I recently came across a website called “The Gender Genie” that claimed to be able to determine a person’s gender from a text they had written. To do this, the site used an algorithm to count how many “masculine” words appeared in the text vs. how many “feminine” words did. I was curious, so I wrote the first few sentences that came to mind. It guessed correctly that I’m female. Then, to see if I could trick the computer, I tried it again intending to sound like a boy. It worked—it guessed that I was male. Here was the breakdown of data:
Analysis |
|||||||||||
Feminine Keywords |
Masculine Keywords |
||||||||||
[with] |
1 |
x |
52 |
= |
52 |
[around] |
0 |
x |
42 |
= |
0 |
[if] |
1 |
x |
47 |
= |
47 |
[what] |
1 |
x |
35 |
= |
35 |
[not] |
0 |
x |
27 |
= |
0 |
[more] |
0 |
x |
34 |
= |
0 |
[where] |
0 |
x |
18 |
= |
0 |
[are] |
0 |
x |
28 |
= |
0 |
[be] |
0 |
x |
17 |
= |
0 |
[as] |
0 |
x |
23 |
= |
0 |
[when] |
0 |
x |
17 |
= |
0 |
[who] |
0 |
x |
19 |
= |
0 |
[your] |
0 |
x |
17 |
= |
0 |
[below] |
0 |
x |
8 |
= |
0 |
[her] |
0 |
x |
9 |
= |
0 |
[is] |
0 |
x |
8 |
= |
0 |
[we] |
0 |
x |
8 |
= |
0 |
[these] |
0 |
x |
8 |
= |
0 |
[should] |
0 |
x |
7 |
= |
0 |
[the] |
4 |
x |
7 |
= |
28 |
[she] |
0 |
x |
6 |
= |
0 |
[a] |
3 |
x |
6 |
= |
18 |
[and] |
4 |
x |
4 |
= |
16 |
[at] |
1 |
x |
6 |
= |
6 |
[me] |
1 |
x |
4 |
= |
4 |
[it] |
5 |
x |
6 |
= |
30 |
[myself] |
0 |
x |
4 |
= |
0 |
[many] |
0 |
x |
6 |
= |
0 |
[hers] |
0 |
x |
3 |
= |
0 |
[said] |
0 |
x |
5 |
= |
0 |
[was] |
3 |
x |
1 |
= |
3 |
[above] |
0 |
x |
4 |
= |
0 |
|
[to] |
4 |
x |
2 |
= |
8 |
For this webpaper, I decided to explore the question of whether it is possible for a computer to determine a person’s gender using nothing but a scientific algorithm.
In our class, we often discussed the idea of blurring boundaries and expanding categories. One of my problems with academia as I have experienced it (excluding this class) is that it is too segmented—Physics is separate from Biology which is separate from English, etc. I have always felt that my areas of interest overlap, and I find it frustrating that I have to major in one thing, being that so many different subjects interest me.
So I thought I would try an experiment – to pose an issue involving gender, information, science, and technology, and approach it from the perspective of three different academic disciplines: psychology, anthropology, and philosophy. These disciplines are the three fields I am most interested in pursuing at college, as I have yet to choose a major. I feel that perhaps by observing how I approach this question from each perspective, I will get a better understanding of what direction I would like to take my education, and maybe even see how much they really overlap.
To do this experiment, I thought it would be helpful to define each field using two different information sources: the Oxford English Dictionary (OED) and Wikipedia. The OED is a scholarly source recognized by academia as being legitimate, while Wikipedia is more culturally relevant, as it is the most widely used encyclopedia on the internet. The table is as follows:
|
Psychology |
Anthropology |
Philosophy |
Oxford English Dictionary |
The scientific study of the nature, functioning, and development of the human mind. |
The science of man, or of mankind, in the widest sense. |
The study of the fundamental nature of knowledge, reality, and existence, and the basis and limits of human understanding. |
Wikipedia Entry |
The science of mind and behavior. |
The study of humanity. |
The study of general and fundamental problems, such as those connected with existence, knowledge, values, reason, mind, and language. |
This table highlights 3 main approaches:
The Self vs. Society vs. The Universe
I will now approach the following question from each perspective to the best of my ability: can a computer determine a person’s gender with a scientific algorithm? What questions would a person who studies this field ask to determine this, and how would they go about exploring them?
*Note: I do not claim to be an expert in any of these areas. This Webpaper is not meant to describe that which is already established by the scientific community to be true, but rather to be an exploration into the unknown. Any facts presented are ones that I have learned collectively over my course of study in all three fields.
Psychology
A psychologist may consider the concept of identity in this problem. In cognitive psychology, gender is considered to be the representation of learned mental patterns. These patterns make up behaviors that, in a social context, often correspond to a gender. Some theories point to hemispheric orientation to explain gender differences, hypothesizing that men think with their “left brain” while women think mostly with their “right brain.” The hemispheres generally focus on the following concepts:
This hypothesis suggests that language and communication would be affected by which hemisphere was dominant. But cognitively speaking, there is a lot of variation between people in general regarding to hemispheric activity in the brain, and we don’t know for sure whether it relates to gender at all.
Gender is a complicated concept and the brain is a complicated organ. A computer would not be able to determine a person’s gender because self-identity can be expressed in many different ways. Gender identity depends on too many factors, both environmental and biological, to be simplified into a random sample of written text. The “Gender Genie” simplifies the gender spectrum into being either male or female. The “feminine” words tend to be words associated with relationships (“with,” “we,” “her”) while the “masculine” words seem more declarative (“are,” “said,” “these”). This test falsely assumes that one’s brain is inherently one or the other.
A psychologist may ask the following questions: how much of a role does personality play in the development of gender identity? What kind of childhood did this person have, and how much do their familial relationships affect their concept of gender? How much does a person’s personality really show in a spontaneously written paragraph? Can someone actually switch mental gender roles in order to trick the computer?
Anthropology
Anthropologists analyze their observations about humanity as a whole. An anthropologist may consider the concept of gender to be largely influenced by their environment. They might criticize the test for its lack of concern for environmental factors—for example, the test does not provide much context for the words in the textbox.
There is no algorithm for predicting the cultural practices of a civilization. There is no formula to determine how gender is represented in any given culture. Therefore, one could not determine such a thing from something so minor as a person’s written words. But they may ask whether it would be more likely for the computer to predict the correct gender for an American person, than for a German or Indian person. Does the test simply reflect cultural beliefs about gender? Would it be more accurate if the person’s cultural identity were part of the algorithm?
How much does our gender affect our social performance? Does it also affect our written performance? Do we “learn” gender through observation of social roles? Can we hide our cultural upbringing at will in order to trick a computer? Has humanity always struggled with the concept of gender and what it should mean to a society?
Philosophy
This is perhaps the most difficult perspective to write from, mostly because it is such a vast topic. It encompasses theories about the entire universe and the nature of reality itself. There are so many branches of philosophy it is almost impossible to approach it in a way that would correspond to so many different systems of thought. Instead, I will propose a potential philosophical response to this question by thinking about it in the broadest, most logical sense.
There is an observable existence of gender in the world. Gender exists, if nothing else, as a concept within the collective consciousness. Thus, it is no wonder that people would be intrigued by a test that can predict it without any contextual clues. The reason why something like this is so fascinating is because gender differences present questions that often cannot be objectively or definitively answered.
One can approach this issue from any number of perspectives, and people have, but perhaps we must accept that there will never be a widely accepted answer. There is no way to know for sure whether gender can be determined by a computation. But I think it’s important to ask these questions not because they need answers, but because we cannot resist. It is an observable fact that we can only advance our society by analyzing and questioning its current condition. We need to be aware of the questions we are asking and understand why we’re asking them, rather than blindly accepting the world as it is. In order to create positive change, we need to be able to question that which we think we already know.
Analysis
As I tried to separate the perspectives, I found it getting more and more difficult. The more I explored each discipline’s view on this question, the more vague the boundaries between them became.
Science is founded upon the pursuit of knowledge and truth through rational analysis of observed phenomena. Each of these disciplines also tries to understand the underlying nature of the world. Rather than examining the physical reality of nature from the micro level (like natural sciences do), the studies of psychology, anthropology, and philosophy seek to understand the foundation of human nature at the macro level. They each focus on examining the complex relationships between humans and their environments.
But no matter how much research has been done in any of these fields, no one has been able to effectively prove a universal truth that explains how technology and gender interact. So perhaps there is no answer to whether a computer can determine a person’s gender from analyzing a paragraph. Maybe sometimes the algorithm will be right, and sometimes it won’t be. Maybe we can’t know for sure what that means about ourselves, or our society, or the workings of the universe. Maybe there just isn’t an answer to this question. And maybe there doesn’t have to be.