This is a script and support imagery for an oral paper presented at the TTU Women’s conference. There is credit due to many and in the written up paper, it will be so. This is simply the 15 minute oral version. But off the top of my head I would have to say I stand on the shoulders, perhaps wobbly because I just climbed up, of James Pennebaker, Robin Lakoff, Deborah Tannen, Molly Ireland, Roderick Hart, and all their cohorts and those who have followed on their heels. Plus Janine Antoni for the lovely performance image from “Loving Care.”
On CERTAINTY – The practice of language of female visual artists’ in social positioning – I THINK. [A computer-based text analysis of language use in female and male artist writings and normative number issues]
As a subset of exploring why female visual artists are still underrepresented in lead exhibition venues, I wondered if female artists performatively ascribe to a social position that may be a factor in perpetuating underrepresentation. Are there linguistic cues in female artists’ writings suggestive of social positioning? In recent psychological and linguistic research findings, the formulation of words in writing and daily practice has been shown to reflect a person’s social and psychological states. After an initial literature review, I turned to the computer-based text analysis tools LIWC-2007 and DICTION-7. Immediately an anomaly arose in the results. The LIWC data correlations, evaluating function words, indicated a strong CERTAINTY component in the language use from the female samples.
This was in direct opposition to the results from DICTION, which evaluates a combination of function and content words. LIWC also indicated a strong certainty component in male artists’ texts, whereas, DICTION found most male samples to fit within in the statistical norm or above. Why the significant discrepancy in the female data? Are female artists less certain than male artist? Is certainty a variable in social positioning? This paper will review my findings generated from female and male artist writing samples through the lens of issues that arose in the quantitative and discourse analysis research practice.
[The statistical analysis and related research necessary for interpreting the data is still in process; hence, more questions than clear avenues of explanation.]
Does vocational success mean I should speak like a man?
As a subset of an inquiry into why female visual artists are still underrepresented in key exhibition venues, I was wondering if gender-linked language use and perception are an additional restrictive factor facing female artists vocationally
Are there linguistic cues in female artists’ writings suggestive of Positioning that are self LIMITING within the art world hierarchy?[/gdlr_quote]
Would a meta-analysis of texts reveal some broad based patterns in this population? PAUSE I don’t actually like Hilary Clinton. Of course, I don’t her, I haven’t researched or followed her, none-the-less, I don’t like her. She is ambitious, direct, unequivocal, intelligent, capable, self-confident and doesn’t take crap. WTF She drags these characteristics around like they are her EIGHT-FOOT PENIS. These features are eviden in her use of language. She positions with gender-linked linguistic power trappings, I associate as masculine. what is wrong with a woman being ambitious, direct, unequivocal, intelligent, capable, and self-confident? ABSOLUTELY fricking nothing. Yet it rubs my internal schema as wrong As masculine. These visceral judments I have had to unpack. Because cognitively I recognize them as wrong. Footnote: I do, however also, respect the hell out of Hillary. She is not underrepresented or set to the side. She is succeeding with her ambitions. YET her linguistic presentation and success begs the question Does vocational success require that she, you or I, speak like a “man?” It is in our lived encounters in which we socially construct the conventions of our Identity and language. This is done typically in relation to power differentials, gradients. Psychosocial linguistic research findings in patterns of vocabularies, structural arrangement, frequency, and thematic content suggest that our texts are reflect our situational psychological + social position. Language is used implicitly + explicitly TO APPRAISE + NAVIGATE POSITIONAL STANCES through cues of affiliation, assertiveness, and achievement. These categories have been further subdivided [via politeness theory] on a spectrum ranging from social infringement to compliance. The cues carry conventions of agency, exchange, judgment, task-orientation, acknowledgment, agreement, support, and decisiveness. And their opposites from the content and structures given in these categorical cues, gaps are filled, interpretations and extrapolations made, and inferences construed to clarify and tweak social hierarchies. Even if the hierarchy is egalitarian in nature. Interestingly, Almost like undetected Freudian slips, pronouns, participles, articles, conjunctions, auxiliary verbs and verb tense are meaningful in the way they correlate with psychological states, personality types, attentional focus, past, future or current mindsets, depression, life stressors, health, deception, status, gender, education, economic status. Content and function words in use are currency in the human pecking order currency. Blah, blah, that was a lot of lists! what did Isay? Let’s look at a super basic example of how language is used to position myself, but your interpretation of my language shifts that positioning according to your own cultural shaping. We POSITION with language…via language choices, arrangement and themes…you read between the lines as you place me in a social pecking order relative to you. Words tend to do three things = Assert +/- Affiliate (even if that means to distance…) +/- Achieve. Not in the above sentence
Remove the function words leaves husband mad…we don’t have a lot of positional context…it is those short lettered words, called function words, that position us that give our sentences social relevance. Let's change a few function words and see how my social positioning begins to shift, crumble. Watch how the changes repositioning me in your mind. Here I have tossed in a little hedge, maybe. I have unfixed the authority of my claim. We could move this little word around in the sentence…changing the meaning Ie my husband said “maybe I am mad about Or what if we swap the function word ABOUT for AT. Ha. That is a pretty big shiftABOUT for AT…AT YOU seriously undermines my position social position. It has changed the affect associated with my affiliation to its opposite. Now….watch how you reframe me as I toss in a short adjective. Not only have I changed my affiliation with my relationship to my husband, I have repositions myself into the divorced category. Bummer on the achievement scale. Despite a liberated mind you may have still shifted your view of me. Even worse, watch how this changes things attitudinally and cues negative data about my affiliation with the former spousal unit. X. I’ve changed the affect from something beign even gentle, former, to a word that is more hostile like the act of scratching his eyes out. XXXXXXXXX. The method of functional word creates positionally packed language categories, used in the everyday, Additionally, the use and understanding of social positioning also show patterns that vary along gender and power status lines. [By gender I do not mean a category defined by vaginas and penises, but by gender-linked cultural and familial practices. Even in this there is evidence to the contrary that status and gender-linked language use and interpretation are fixed.
- MY = mine, ownership
- HUSBAND = pronouncing belonging | and hints at my heterosexuality
- SAID = claim
- I am MAD ABOUT you = through affect, location, possession, positional | bolstering my positionYet language is inherently social and positions us.
My question again, in the written language of female visual artists are there patterns of positioning that suggest they may LIMIT themselves within the art world hierarchy? This question is loaded and I have yet to unpack all the variables and sub-variables for analysis. In the future I am sure I will have to whittle this down to a series of subquestions that I might actually be able to track down tendencies that might be relevant.
After an initial review of gender-linked language patterns, politeness theory, power and dominance, having introducing myself to seminal texts by researchers, Robin Lakoff, Deborah Tannen, James Pennebaker and cohorts, along with subsequent studies that both confirm and call into question their suppositions of…power | male linked tendencies – convey information, direct, succinct, instrumental (task/object oriented)
And, female linked language tendencies – convey connections, indirect, elaborative, affective, self-reference.
I turned to psychosocial linguistic computer-based text analysis tools, LIWC-2007 and DICTION-7, to attempt to partially begin uncovering tendencies, if they existed. They are tools I have just jumped off the deep end in to–meaning more research on methods and tool usage may shift my reading of my data. So this is my first stab, jump, fall, whatever.
Oh, another problem I have run into is that most frameworks in understanding language specifically around power or gender are like models for light. Light has behavioral tendencies like a wave form, but look at it from a different frame and it behaves like a particle. Which model is correct? So frameworks and models are just conveniences to navigate interpretations of highly complex events or practices.They help with pattern comprehension.
LIWC-2007, “Linguistic Inquiry and Word Count is a text analysis software program developed by James Pennebaker, Roger Booth, and Martha Francis. LIWC calculates the degree to which people use different categories of words and word combinations across a wide array of text genres, including emails, speeches, poems, blogs, twitter, transcribed daily speech, and responsive writings acquired as a part of empirical studies.”
LIWC relies on word count, pattern analysis and thematic tendencies found in texts. It creation was for the intent of being a tool for corpus analysis to identify inter-subjective patterns. It is a quantitative tool that used traditional qualitative practices to verify each aspect of its development. Normative* patterns from 400,000 text files have emerged from LIWC, across seven plus written and spoken genres. These statistical norms are not fully fixed across each genre or situation, but they indicate various tendencies.
*Normative is relative and should alway be taken with at least two grains of salt.
The second analytic software tool I am using is DICTION-7 developed by Roderick Hart and cohorts. Like LIWC, it uses word counts, phrase analysis, and to a greater degree theme evident in content words. It is heavily geared toward research in politics and focuses on measuring language tone. It characterizes texts according to five master variables—activity, optimism, certainty, realism, and commonality. Each master variable is composed of categories of linguistic sub-features.
Its normative numbers for the linguistic sub-features have largely been established by samplings of writings and transcripts from politicians (presidents, judges, congressmen, etc) and journalists for 50 plus years, including current. I assumed this meant the normatives were harvested predominantly from men. I shot an email to Hart to confirm or counter this; his team (wife) confirmed it.
So I collected 159 artist written texts. Though Random selection is a nice thought when trying to analyze population, there is really no such thing. I chose my the visual artist texts based on several conditions –
(1) the artists were vetted, recognized in scholarly literature, as artists by art theorists and historians, specifically I chose artists included in Kristine Stiles and Peter Selz’s anthology of artists’ writings, Theories and Documents of Contemporary Art: A Sourcebook of Artists’ Writings, 2nd Ed;
(2) Texts had to be readily accessible online (no typing); and
(3) relatively contemporary, working within the last 65 years.
Further, I set out to draw
(4) an equal number from each decade;
(5) limited individual authors to 3 essays;
(6) varied text genre including manifestos, letters, poems, artist statements, journal entries, published essays and book excerpts.
Each text was cleaned up for spelling errors and non-computing contractions and identified by artist name, biological gender, date authored, and genre.
Break out your salt shakers and take a few grains! [hedge]
Though I am not yet formally trained in statistics, as artist, it is part of my creative practice to identify patterns and anomalies. Even as untrained statistician, these were easily evident in the computer analysis of my sample texts. Most notably were those along variables associated with gender-linked language and CERTAINTY. Compared to the general public norms identified by LIWC, the female artists tended toward male-linked language practices and were generally high on LIWCs CERTAINTY variable. Yet, compared to the normatives of DICTION (recall these are composed of male politicians and journalists) the artists’ gender disparity roughly fell in line with male/female patterns presented by Lakoff in 1975. at least relationally and, there were significant anomalies evident in regards to the master variable of CERTAINTY.
NOTE in LIWC the female artist patterns, still leaning toward the masculine practices, yet registered higher than the male artists, in these feminine practices of words per sentence, self-reference (I/me/my), social, and affect.
In male language practices, the female artists registered with higher frequency then the male artists in achievement, swear words, and quantifying statements.
The top five linguist variables, in terms of percentages of the number of essays that sat ABOVE DICTIONS norms for female artists included INSISTENCE, SELF-REFERENCE, VARIETY, CERTAINTY, COGNITION. I have bolded the masculine linked practices. and underlined the more feminine practices. The blue are some of the males top 5. Note the differences. So 40 and 60% of artist essays were more insistent than journalisitic and political norms. Self-reference makes sense based on the fact that artists are speaking to their personal work not broad communities. That so many essays registered above DICTIONs norms is interesting and suggestive of arena’s of further research.
Here are the top 5 variables in which essays fellow BELOW dictions norms COLLECTIVES, INSISTENCE, (40/60 where above…leaving very few relative to this variable within the norm), CERTAINTY, (again certainty showed up at high frequency above.) PRAISE, CONCRETENESS. (being low on this variable just means they were addressing more conceptual notions).
In DICTION CERTAINTY is built by adding the %
subcategories linguistic markers of + tenacity [is, am, will, shall, has, must, do] + leveling [ignore difference] + collective [crowd, army, team] + insistence [words used 3+, noun, derivatives], and SUBTRACTING – numerical [hyperspecificity] – ambivalence [hesitancy, unwillingness to commit – self-referential [first person] – variety[avoidance of overstatement].
LETS LOOK MORE CLOSELY AT the sub variables that compose CERTAINTY. Here are the numbers of essays as they correlated with dictions male NORMS. At the very left I have color coded linguistic markers that correlate to masculine or feminine stereotyped practices-blue/pink. Then listing the sub variables measured the top are additive, the bottom are subtracted in calculating the degree of linguistic certainty expressed in a text. NOTE Insistence, in yellow, HOW few essays for both male/female artists are in the normative range
Again note INSISTENCE in yellow…these are the number of essays that ranked above the norms of politicians and journalists. Note the difference between males and females. Male artists are damn insistent…does this impact their higher degree of representation ??? It is an area strongly suggestive of further exploration!
Here are the essays that fell below dictions norms. Collective actually makes sense based on artists practices being relatively isolated. Again here we see insistence…based on essays above and below … very few artists essays fell within dictions insistence NORMS.
This definitely peak questions?
Can I answer my initial question. NOT really. From lived experience, we use language everyday, to establish ours and others social positions. But I cannot address from my current data set whether there are traces that hint at female artists limiting themselves. Are there gender practices, yes. How they correlate to power and vocational reception in the field of art is yet to be determined. I have a lot more research to begin to understand their meaning.
PROBLEMS I am butting my head against with my current data/research set.
SAMPLES: 1.) set of female artists I have chosen are recognized/represented; 2.) I’ve categorized by biology sex not ascribed practicing community; 3.) all the texts were written in social isolation versus direct social engagement; 4.) I have not compared along analyze by particular genre practice (painting verus sculpture) nor whether the texts are diary entries, art essays, or manifestos; and 5.) how much the texts was edited by others is unknown.
RESEARCH: 1.) messiness of literature available on gender versus power difference, its contextual conflictedness/shiftiness; 2.) measurability of power/gender privileged?
RESEARCHER: 1.) lack of background in statistics; 2.) neophyte with tool LIWC + DICTION; 3.) lack of domain depth in gender, language + power
SAMPLES: collect and analyze interview samples, gender identification vs biological sex, establish artist norms + professional equivalents
RESEARCH: more HW gender versus power, vocations that privilege male speech
RESEARCHER: learn statistics, learn tool LIWC + DICTION, learn of domain depth in gender, language + power
I am sTYMIED with the question. Does vocational success for the female artist mean speak like a man?Can this even be measured.
The door is open for a lot more research to explore items within the current texts: gender choice versus sexual biological — language differences. Further directions include anaylsis interview texts to see if positioning is more evident.
Tune in later for more. I take my first research methods/language class this summer.