INVESTIGATING GOOGLE’S SEARCH ENGINE Bloomsbury Studies in Digital Cultures Series Editors Anthony Mandal and Jenny Kidd This series responds to a rapidly changing digital world, one which permeates both our everyday lives and the broader philosophical challenges that accrue in its wake. It is inter- and trans-disciplinary, situated at the meeting points of the digital humanities, digital media and cultural studies, and research into digital ethics. While the series will tackle the ‘digital humanities’ in its broadest sense, its ambition is to broaden focus beyond areas typically associated with the digital humanities to encompass a range of approaches to the digital, whether these be digital humanities, digital media studies or digital arts practice. Titles in the series The Trouble With Big Data, Jennifer Edmond, Nicola Horsley, Jörg Lehmann and Mike Priddy Queer Data, Kevin Guyan Hacking in the Humanities, Aaron Mauro Forthcoming titles Human Exploits, Cyberpunk and the Digital Humanities, Aaron Mauro Ambient Stories in Practice and Research, Edited by Amy Spencer Metamodernism and the Postdigital in the Contemporary Novel, Spencer Jordan Herman Melville and the Digital Humanities, Christopher Ohge and Dennis Mischke Listening In, Toby Heys, David Jackson and Marsha Courneya People Like You, Sophie Day, Celia Lury and Helen Ward INVESTIGATING GOOGLE’S SEARCH ENGINE ETHICS, ALGORITHMS, AND THE MACHINES BUILT TO READ US Rosie Graham BLOOMSBURY ACADEMIC Bloomsbury Publishing Plc 50 Bedford Square, London, WC1B 3DP, UK 1385 Broadway, New York, NY 10018, USA 29 Earlsfort Terrace, Dublin 2, Ireland BLOOMSBURY, BLOOMSBURY ACADEMIC and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2023 Copyright © Rosie Graham, 2023 Rosie Graham has asserted her right under the Copyright, Designs and Patents Act, 1988, to be identified as Author of this work. For legal purposes the Acknowledgements on p. xiv constitute an extension of this copyright page. Cover design by Rebecca Heselton Cover image © J614/ Getty Images All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. ISBN: HB: 978-1-3503-2520-3 PB: 978-1-3503-2519-7 ePDF: 978-1-3503-2521-0 eBook: 978-1-3503-2522-7 Typeset by Deanta Global Publishing Services, Chennai, India Series: Bloomsbury Studies in Digital Cultures To find out more about our authors and books visit www .bloomsbury .com and sign up for our newsletters. This book is dedicated to: Regenia Gagnier for your fierce intellectualism and encouragement. Anne and Keith for your love, support, and patience. Zoe Hope Bulaitis for everything and more. vi CONTENTS List of illustrations x Acknowledgements xiv Introduction: Investigating Google’s search engine 1 Google’s dominance 2 The three steps of how search engines work: Crawling, ranking, and query results 4 Step one: Crawling 4 Step two: Ranking 4 Step three: Query results 8 Five key challenges of studying Google’s search engine 10 One: Multiple actors: Search engine optimization and economic incentives 11 Two: Moving targets 11 Three: Each search a partial viewpoint 12 Four: No real alternatives 13 Five: The myth of black boxes 14 Chapter outlines 15 Chapter 1: Understanding Google queries and the problem of intentions 16 Chapter 2: Google’s impact on cognition and memory: Histories, concepts, and technosocial practices 16 Chapter 3: Autocomplete: Stereotypes, biases, and designed discrimination 17 Chapter 4: Google’s search engine results: What is a relevant result? 17 Chapter 5: The real cost of search engines: Digital advertising, linguistic capitalism, and the rise of fake news 18 Notation and examples 18 1 Understanding Google queries and the problem of intentions 21 Introduction 21 Categorizing how and what people search 23 The roles of search engines and information retrieval’s question of why 24 Query length and the problems of intention 28 All information is ethical: Searching for [food for snakes] 30 Predicting intentions with a lack of information: Plato, Gadamer, and Derrida 32 Gadamer’s hermeneutics and Plato’s fears of deception 35 Google’s algorithms and Derrida’s monster 38 Contents What kinds of things do people search Google for? 40 Google trends, Brexit, and ‘frantically’ googling after the EU referendum 42 Conclusion 47 2 Google’s impact on cognition and memory: Histories, concepts, and technosocial practices 49 Introduction 49 Google’s impact on cognition and memory 50 Kinds of recall from extended minds to transactive memory 54 Technosocial memory practices from oral culture to digital literacy 57 The legacy of naturalized technologies 60 Truth and knowledge for Plato 62 Aristotle’s sensory approach 64 Technosocial memory before Google: The Ars Memoria 65 The science and magic of search 68 Treating the mind as technology: Bacon, Hooke, and modern psychology 73 Conclusion 77 3 Autocomplete: Stereotypes, biases, and designed discrimination 79 Introduction 79 The desire for a digital oracle 82 Autocomplete’s minimal academic attention 85 The biases of Autocomplete: Stereotypes and discrimination 88 Predicting and shaping user attitudes: The origins of Autocomplete 93 So, how does Autocomplete operate? 95 Second-order stereotyping: Sexist suggestions for female scientists 97 RankBrain and the biases of machine learning 104 Automated misogyny for every individual 113 Speed 116 Speed and judgement: Time to reflect 120 Conclusion 126 4 Google’s search engine results: What is a relevant result? 129 Introduction 129 ‘Quantifiable signals’ and Malawian witch doctors 131 What should search engine results be? 132 The Idealists: Search is democratic, relevance can be measured objectively, and answers can exist independently of bias 134 The difficulty with measuring relevance 136 The Contextualists: Search is undemocratic, relevance is a measure of personalization, and all answers are inherently biased 139 Are search results personalized? 141 viii Contents Methodological challenges of studying search engines 143 Particular considerations for collecting search engine results 144 Variables that matter: Search experiments in 2015, 2017, and 2021 146 The rationale behind focusing on same-sex sexual orientation 146 Queries used 148 Capturing the spread of results from the first page 149 Evaluation method 150 Google’s public position on how they provide results 150 Summary of 2015 results 155 How do variations in terminology and phrasing alter search results? 161 Unimaginable communities 166 How search results change throughout time: 2015, 2017, and 2021 167 Longitudinal overview: Official languages in each domain 168 Terminology throughout time: ‘Homosexual’ versus ‘gay’ 173 Phrasing throughout time: ‘Good’ versus ‘wrong’ 177 Conclusion 179 5 The real cost of search engines: Digital advertising, linguistic capitalism, and the rise of fake news 181 Introduction 181 The economics of google 182 The context of post-Fordism 184 AdWords: Organic versus sponsored results 185 AdWords: The multilingual linguistic market and an economy of bias 187 Google’s institutionalization, data collection, and advertising 191 AdWords in the context of ‘The Magic System’ 193 AdWords and the general intellect 195 The economic profits of discrimination 197 Private profits and public losses 200 Google’s international expansion 202 AdSense and post-Fordism: The cost of Google’s billboards 203 AdSense and fake news in the 2016 US presidential election 204 The reciprocal relationship between AdSense and Facebook 207 Conclusion 211 Conclusion: What if search engines were actually built to benefit users? 213 Bibliography 217 Index 234 ix