Minor Moves: Growth, Fugitivity, and Children's Physical Movement

From tendencies to reduce the Underground Railroad to the imperative "follow the north star" to the iconic images of Ruby Bridges' 1960 "step forward" on the stairs of William Frantz Elementary School, America prefers to picture freedom as an upwardly mobile development. This preoccupation with the subtractive and linear force of development makes it hard to hear the palpable steps of so many truant children marching in the Movement and renders illegible the nonlinear movements of minors in the Underground. Yet a black fugitive hugging a tree, a white boy walking alone in a field, or even pieces of a discarded raft floating downstream like remnants of child's play are constitutive gestures of the Underground's networks of care and escape. Responding to 19th-century Americanists and cultural studies scholars' important illumination of the child as central to national narratives of development and freedom, "Minor Moves" reads major literary narratives not for the child and development but for the fugitive trace of minor and growth.

In four chapters, I trace the physical gestures of Nathaniel Hawthorne's Pearl, Harriet Beecher Stowe's Topsy, Harriet Wilson's Frado, and Mark Twain's Huck against the historical backdrop of the Fugitive Slave Act and the passing of the first compulsory education bills that made truancy illegal. I ask how, within a discourse of independence that fails to imagine any serious movements in the minor, we might understand the depictions of moving children as interrupting a U.S. preoccupation with normative development and recognize in them the emergence of an alternative imaginary. To attend to the movement of the minor is to attend to what the discursive order of a development-centered imaginary deems inconsequential and what its grammar can render only as mistakes. Engaging the insights of performance studies, I regard what these narratives depict as childish missteps (Topsy's spins, Frado's climbing the roof) as dances that trouble the narrative's discursive order. At the same time, drawing upon the observations of black studies and literary theory, I take note of the pressure these "minor moves" put on the literal grammar of the text (Stowe's run-on sentences and Hawthorne's shaky subject-verb agreements). I regard these ungrammatical moves as poetic ruptures from which emerges an alternative and prior force of the imaginary at work in these narratives--a force I call "growth."

Reading these "minor moves" holds open the possibility of thinking about a generative association between blackness and childishness, one that neither supports racist ideas of biological inferiority nor mandates in the name of political uplift the subsequent repudiation of childishness. I argue that recognizing the fugitive force of growth indicated in the interplay between the conceptual and grammatical disjunctures of these minor moves opens a deeper understanding of agency and dependency that exceeds notions of arrested development and social death. For once we interrupt the desire to picture development (which is to say the desire to picture), dependency is no longer a state (of social death or arrested development) of what does not belong, but rather it is what Édouard Glissant might have called a "departure" (from "be[ing] a single being"). Topsy's hard-to-see pick-pocketing and Pearl's running amok with brown men in the market are not moves out of dependency but indeed social turns (a dance) by way of dependency. Dependent, moving and ungrammatical, the growth evidenced in these childish ruptures enables different stories about slavery, freedom, and childishness--ones that do not necessitate a repudiation of childishness in the name of freedom, but recognize in such minor moves a fugitive way out.

Facts and Fictions: Feminist Literary Criticism and Cultural Critique, 1968-2012

"Facts and Fictions: Feminist Literary Criticism and Cultural Critique, 1968-2012" is a critical history of the unfolding of feminist literary study in the US academy. It contributes to current scholarly efforts to revisit the 1970s by reconsidering often-repeated narratives about the critical naivety of feminist literary criticism in its initial articulation. As the story now goes, many of the most prominent feminist thinkers of the period engaged in unsophisticated literary analysis by conflating lived social reality with textual representation when they read works of literature as documentary evidence of real life. As a result, the work of these "bad critics," particularly Kate Millett and Andrea Dworkin, has not been fully accounted for in literary critical terms.

This dissertation returns to Dworkin and Millett's work to argue for a different history of feminist literary criticism. Rather than dismiss their work for its conflation of fact and fiction, I pay attention to the complexity at the heart of it, yielding a new perspective on the history and persistence of the struggle to use literary texts for feminist political ends. Dworkin and Millett established the centrality of reality and representation to the feminist canon debates of "the long 1970s," the sex wars of the 1980s, and the more recent feminist turn to memoir. I read these productive periods in feminist literary criticism from 1968 to 2012 through their varied commitment to literary works.

Chapter One begins with Millett, who de-aestheticized male-authored texts to treat patriarchal literature in relation to culture and ideology. Her mode of literary interpretation was so far afield from the established methods of New Criticism that she was not understood as a literary critic. She was repudiated in the feminist literary criticism that followed her and sought sympathetic methods for reading women's writing. In that decade, the subject of Chapter Two, feminist literary critics began to judge texts on the basis of their ability to accurately depict the reality of women's experiences.

Their vision of the relationship between life and fiction shaped arguments about pornography during the sex wars of the 1980s, the subject of Chapter Three. In this context, Dworkin was feminism's "bad critic." I focus on the literary critical elements of Dworkin's theories of pornographic representation and align her with Millett as a miscategorized literary critic. In the decades following the sex wars, many of the key feminist literary critics of the founding generation (including Dworkin, Jane Gallop, Carolyn Heilbrun, and Millett) wrote memoirs that recounted, largely in experiential terms, the history this dissertation examines. Chapter Four considers the story these memoirists told about the rise and fall of feminist literary criticism. I close with an epilogue on the place of literature in a feminist critical enterprise that has shifted toward privileging theory.

Bayesian Inference in Large-scale Problems

Many modern applications fall into the category of "large-scale" statistical problems, in which both the number of observations n and the number of features or parameters p may be large. Many existing methods focus on point estimation, despite the continued relevance of uncertainty quantification in the sciences, where the number of parameters to estimate often exceeds the sample size, despite huge increases in the value of n typically seen in many fields. Thus, the tendency in some areas of industry to dispense with traditional statistical analysis on the basis that "n=all" is of little relevance outside of certain narrow applications. The main result of the Big Data revolution in most fields has instead been to make computation much harder without reducing the importance of uncertainty quantification. Bayesian methods excel at uncertainty quantification, but often scale poorly relative to alternatives. This conflict between the statistical advantages of Bayesian procedures and their substantial computational disadvantages is perhaps the greatest challenge facing modern Bayesian statistics, and is the primary motivation for the work presented here.

Two general strategies for scaling Bayesian inference are considered. The first is the development of methods that lend themselves to faster computation, and the second is design and characterization of computational algorithms that scale better in n or p. In the first instance, the focus is on joint inference outside of the standard problem of multivariate continuous data that has been a major focus of previous theoretical work in this area. In the second area, we pursue strategies for improving the speed of Markov chain Monte Carlo algorithms, and characterizing their performance in large-scale settings. Throughout, the focus is on rigorous theoretical evaluation combined with empirical demonstrations of performance and concordance with the theory.

One topic we consider is modeling the joint distribution of multivariate categorical data, often summarized in a contingency table. Contingency table analysis routinely relies on log-linear models, with latent structure analysis providing a common alternative. Latent structure models lead to a reduced rank tensor factorization of the probability mass function for multivariate categorical data, while log-linear models achieve dimensionality reduction through sparsity. Little is known about the relationship between these notions of dimensionality reduction in the two paradigms. In Chapter 2, we derive several results relating the support of a log-linear model to nonnegative ranks of the associated probability tensor. Motivated by these findings, we propose a new collapsed Tucker class of tensor decompositions, which bridge existing PARAFAC and Tucker decompositions, providing a more flexible framework for parsimoniously characterizing multivariate categorical data. Taking a Bayesian approach to inference, we illustrate empirical advantages of the new decompositions.

Latent class models for the joint distribution of multivariate categorical, such as the PARAFAC decomposition, data play an important role in the analysis of population structure. In this context, the number of latent classes is interpreted as the number of genetically distinct subpopulations of an organism, an important factor in the analysis of evolutionary processes and conservation status. Existing methods focus on point estimates of the number of subpopulations, and lack robust uncertainty quantification. Moreover, whether the number of latent classes in these models is even an identified parameter is an open question. In Chapter 3, we show that when the model is properly specified, the correct number of subpopulations can be recovered almost surely. We then propose an alternative method for estimating the number of latent subpopulations that provides good quantification of uncertainty, and provide a simple procedure for verifying that the proposed method is consistent for the number of subpopulations. The performance of the model in estimating the number of subpopulations and other common population structure inference problems is assessed in simulations and a real data application.

In contingency table analysis, sparse data is frequently encountered for even modest numbers of variables, resulting in non-existence of maximum likelihood estimates. A common solution is to obtain regularized estimates of the parameters of a log-linear model. Bayesian methods provide a coherent approach to regularization, but are often computationally intensive. Conjugate priors ease computational demands, but the conjugate Diaconis--Ylvisaker priors for the parameters of log-linear models do not give rise to closed form credible regions, complicating posterior inference. In Chapter 4 we derive the optimal Gaussian approximation to the posterior for log-linear models with Diaconis--Ylvisaker priors, and provide convergence rate and finite-sample bounds for the Kullback-Leibler divergence between the exact posterior and the optimal Gaussian approximation. We demonstrate empirically in simulations and a real data application that the approximation is highly accurate, even in relatively small samples. The proposed approximation provides a computationally scalable and principled approach to regularized estimation and approximate Bayesian inference for log-linear models.

Another challenging and somewhat non-standard joint modeling problem is inference on tail dependence in stochastic processes. In applications where extreme dependence is of interest, data are almost always time-indexed. Existing methods for inference and modeling in this setting often cluster extreme events or choose window sizes with the goal of preserving temporal information. In Chapter 5, we propose an alternative paradigm for inference on tail dependence in stochastic processes with arbitrary temporal dependence structure in the extremes, based on the idea that the information on strength of tail dependence and the temporal structure in this dependence are both encoded in waiting times between exceedances of high thresholds. We construct a class of time-indexed stochastic processes with tail dependence obtained by endowing the support points in de Haan's spectral representation of max-stable processes with velocities and lifetimes. We extend Smith's model to these max-stable velocity processes and obtain the distribution of waiting times between extreme events at multiple locations. Motivated by this result, a new definition of tail dependence is proposed that is a function of the distribution of waiting times between threshold exceedances, and an inferential framework is constructed for estimating the strength of extremal dependence and quantifying uncertainty in this paradigm. The method is applied to climatological, financial, and electrophysiology data.

The remainder of this thesis focuses on posterior computation by Markov chain Monte Carlo. The Markov Chain Monte Carlo method is the dominant paradigm for posterior computation in Bayesian analysis. It has long been common to control computation time by making approximations to the Markov transition kernel. Comparatively little attention has been paid to convergence and estimation error in these approximating Markov Chains. In Chapter 6, we propose a framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified loss function and computational budget. The results require only ergodicity of the exact kernel and control of the kernel approximation accuracy. The theoretical framework is applied to approximations based on random subsets of data, low-rank approximations of Gaussian processes, and a novel approximating Markov chain for discrete mixture models.

Data augmentation Gibbs samplers are arguably the most popular class of algorithm for approximately sampling from the posterior distribution for the parameters of generalized linear models. The truncated Normal and Polya-Gamma data augmentation samplers are standard examples for probit and logit links, respectively. Motivated by an important problem in quantitative advertising, in Chapter 7 we consider the application of these algorithms to modeling rare events. We show that when the sample size is large but the observed number of successes is small, these data augmentation samplers mix very slowly, with a spectral gap that converges to zero at a rate at least proportional to the reciprocal of the square root of the sample size up to a log factor. In simulation studies, moderate sample sizes result in high autocorrelations and small effective sample sizes. Similar empirical results are observed for related data augmentation samplers for multinomial logit and probit models. When applied to a real quantitative advertising dataset, the data augmentation samplers mix very poorly. Conversely, Hamiltonian Monte Carlo and a type of independence chain Metropolis algorithm show good mixing on the same dataset.