Seepage, objectivity, and climate science

Seepage, objectivity, and climate science

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx Contents lists available at ScienceDirect Studies in History and Philosophy of Scien...

1MB Sizes 0 Downloads 4 Views

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Studies in History and Philosophy of Science journal homepage: www.elsevier.com/locate/shpsa

Seepage, objectivity, and climate science Ryan O'Loughlin Department of History and Philosophy of Science and Medicine, Indiana University-Bloomington, 1165 E. 3rd St. Morrison Hall Rm. 302, Bloomington, IN, 47405, USA

HIGHLIGHTS

is a useful concept for assessing the impact of climate contrarianism on the climate science community • ·Objectivity about climate processes gained from studying the "hiatus" is legitimate, despite seepage having occurred • Knowledge global warming "hiatus" case involves successes and partial failures of objectivity in the climate science community • The • The disproportionate amount of focus on the "hiatus" time period may partially be a product of “hobby-horse” riding ARTICLE INFO

ABSTRACT

Keywords: Seepage Global warming hiatus Global warming slowdown Objectivity Climate science communication

Based on the disproportionate amount of attention paid by climate scientists to the supposed global warming hiatus, it has recently been argued that contrarian discourse has “seeped into” climate science. While I agree that seepage has occurred, its effects remain unclear. This lack of clarity may give the impression that climate science has been compromised in a way that it hasn't—such a conclusion should be defended against. To do this I argue that the effects of seepage should be analyzed in terms of objectivity. I use seven meanings of objectivity to analyze contrarian discourse's impact on climate science. The resulting account supports the important point that climate science has not been compromised in a way that invalidates the conclusions its scientists have drawn, despite the reality of seepage having occurred.

1. Introduction In this paper I discuss the global warming “hiatus”, which was an apparent pause in the rate of increase in global mean surface temperature (GMST) that occurred from 1998 to 2012. More specifically I discuss the relationship between climate contrarian discourse that focused on the “hiatus” and how this discourse impacted the climate science community.1 Based on the disproportionate amount of attention paid to the “hiatus” by climate scientists, it has recently been argued by Stephan Lewandowsky, Naomi Oreskes, James Risbey, Ben Newell, and Michael Smithson (2015) that contrarian discourse has “seeped into” climate science (see section 2). While I agree with these authors that seepage has occurred, its effects are unclear based on their analysis, as they focus on the cause (and existence) of seepage. Moreover, this lack

of clarity may give the impression that climate science has been compromised in a way that it has not—such a conclusion should be defended against. In order to mount such a defense, I argue that the effects of seepage should be understood in terms of objectivity. As Heather Douglas (2004) and Elisabeth Lloyd & Vanessa Schweizer (2014) demonstrate, objectivity can be understood to have several meanings (being detached when assessing evidence, the results of a method being unbiased in a statistical sense, a procedure being replicable, as a few examples). I will use the meanings discussed by Lloyd & Schweizer (2014) to provide a detailed account of contrarian discourse's impact on climate science.2 I begin, in Section 2, by giving some background information on the “hiatus” controversy and on Lewandowsky et al.’s assessment. Section 3 adds to Lewandowsky et al.’s account by highlighting the importance of

E-mail address: [email protected] I use “contrarian” throughout this paper, rather than “denialist” or “skeptic”. “Denialism” could refer to the denial of several different claims associated with climate change (e.g., denial that “the Earth has been warming”, or that “the Earth has been warming due to human causes”, or that “there is a scientific consensus concerning the Earth's warming”, etc.). On the other hand, “skeptic” has a positive connotation and so should also be avoided. Following McCright (2007) and O'Neill and Boykoff (2010), I use “contrarian” to refer to those who “vocally challenge what they see as a false consensus of mainstream climate science through critical attacks on climate science and eminent climate scientists, often with substantial financial support from fossil fuels industry organizations and conservative think tanks” (O'Neill and Boykoff 2010, E151). 2 Lloyd and Schweizer (2014) use the meanings of objectivity analyzed by Lloyd (1995) along with two additional (social) meanings. 1

https://doi.org/10.1016/j.shpsa.2019.07.002 Received 13 March 2019; Received in revised form 26 June 2019; Accepted 15 July 2019 0039-3681/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Ryan O'Loughlin, Studies in History and Philosophy of Science, https://doi.org/10.1016/j.shpsa.2019.07.002

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

both hobby-horse riding and the role climate scientists have as communicators. Finally, section 4 evaluates the damage done to climate science by the “hiatus” controversy using the various meanings of objectivity.

note that the IPCC adopted “the term “hiatus” in its Fifth Assessment Report” despite its inaccuracy and despite a specific comment advising against the use of the term by the German government (2015, 6–7). Additionally, see Fig. 1, taken from Medhaug, Stolpe, Fischer, and Knutti (2017), which shows the trends in climate science publications and Google search queries on the “global warming pause” and “global warming hiatus”. In total, there is ample evidence that climate scientists paid a disproportionate amount of attention to the “hiatus”.

2. The global warming “hiatus” and seepage Despite the consensus among climate scientists that the earth has been and is still warming primarily due to anthropogenic greenhouse gas emissions, the global mean surface temperature (GMST) didn't appear to rise from 1998 to 2012. The reasons for this apparent slowdown in warming are manifold; two important reasons include the cherrypicked data of a very strong El Niño event in 1998 leading to an anomalously warm GMST for that year and some problems with the early observational data.3 While the apparent slowdown is not statistically significant, this didn't stop climate contrarians from focusing on it to call climate models into question, claiming both that global warming had stopped and that it has gone on hiatus, thereby creating a “hiatus” framing.4 Indeed, as Lewandowsky et al. note, “the media and other public actors created a frame for the discussion of climate change that focused on the allegation that global warming … [had] entered a “hiatus”” (2015, 5).5 This is significant because according to Lewandowsky et al., referring to the time period as a “hiatus” implies both “that the period in question … is more than a routine short-term fluctuation or deviation from a long-term trend” and that climate scientists have misunderstood certain important physical climate processes (2015, 6). Indeed, the google trends search criteria (see Fig. 1) shows an increasing public interest in the global warming “hiatus” itself, and an article in Mother Jones, an American Magazine, details the overblown way in which the media treated the “hiatus” directly before the publication of the IPCC (the UN's Intergovernmental Panel on Climate Change)'s Fifth Assessment Report (Mooney, 2013). Thus, it is evident that the apparent “hiatus” impacted the media/public's view of climate science and of climate models, due in large part to contrarian discourse. Lewandowsky et al. take things a step further arguing that climate contrarian discourse “seeped into” the work of climate scientists themselves in this particular episode. These authors define “seepage” as the “infiltration and influence of … essentially non-scientific claims into scientific work” and ultimately claim that as a result of contrarian discourse, “In effect, scientists came to doubt their own conclusions, and felt compelled to do more work to further strengthen them, even if this meant discarding previously accepted standards of statistical practice” (2015, 9). They support this claim by discussing certain cognitive processes that could give rise to seepage and claim that climate scientists may be susceptible to the influences of contrarian discourse based on cognitive research that examines these processes.6 Lewandowsky et al. further argue that seepage has actually occurred by citing the disproportionately large amount of attention paid to the “hiatus” by climate scientists in their research. For instance, there were “two 2014 special issues/sections of Nature journals devoted to the “pause” or “hiatus” (Nature Climate Change, March 2014, No. 149; and Nature Geoscience, February 2014, No. 157)” and Lewandowsky et al.

3. The seepage analysis The above evidence discussed by Lewandowsky et al. makes it clear that seepage has occurred in some sense. Indeed, for climate scientists to focus so heavily on such a small time period, of 15 years, despite its lack of statistical significance demonstrates that extra-scientific factors—including contrarian discourse—likely had an impact on the scientific community. However, it is not clear in what sense climate scientists adopted the “hiatus” framing put forth by contrarians. This is partially due to the fact that framings are oftentimes not explicitly stated or defined. Is it correct to understand the global warming “hiatus” framing as follows?: “hiatus” framing: the GMST slowing/stopping trend from 1998-2012 shows that there is something seriously wrong with climate models, especially the role of anthropogenic GHG emissions in the models. Lewandowsky, Risbey, and Oreskes (2016) address this issue by defining the framing, which will be discussed below, but I will also suggest that there is a hobby-horse explanation that is relevant to the framing issue that has been overlooked. Additionally, the role of climate scientists as communicators must be emphasized to fully understand how and why seepage occurred with this particular scientific episode. Both the hobby-horse explanation and the point about climate scientists being communicators are complementary to Lewandowsky et al.’s seepage analysis in terms of understanding why climate scientists paid so much attention to the “hiatus” in their research. Moreover, both are fruitful for assessing how seepage has impacted the objectivity of climate science (section 4). 3.1. The “hiatus” framing issue and the hobby-horse explanation Whether climate scientists adopted the “hiatus” framing depends on what one means by “framing”.7 Indeed, the definition of the “hiatus” framing given at the beginning of section 3 is much more explicit and detailed than framings typically are when they are used. This is pointed out by Lewandowsky, Risbey, and Oreskes in another paper, “The ‘Pause’ in Global Warming: Turning a Routine Fluctuation into a Problem for Science” (2016). They state that “Frames are rhetorical and communicative structures that select and highlight certain aspects of a perceived reality over others (Dirikx & Gelders, 2010)”, noting that “frames are rarely made explicit” (Lewandowsky et al. 2016, 728). Based on this notion of a “pause” or “hiatus” framing, and based on their analysis of the abstracts and opening paragraphs of various climate science articles that discuss the “pause” or “hiatus”, Lewandowsky, Risbey, and Oreskes conclude that,

3 See Medhaug et al. (2017) for a discussion of these and other reasons for the apparent “hiatus” in warming. 4 I use quotes around “hiatus” throughout the paper to remind readers that there was not a statistically significant hiatus in warming. 5 Lewandowsky et al. also note that “pause” or “stall” (and other related words) are sometimes used when talking about the “hiatus”. In this paper, I use “hiatus”, and I take it that Lewandowsky et al.’s claims about the “pause framing” apply equally well to a “hiatus framing” since they often interchangeably use those terms. 6 Cognitive processes examined include stereotype threat (Brysse, Oreskes, O’Reilly, & Oppenheimer, 2013; Freudenburg and Muselli, 2010), pluralistic ignorance (Leviston, Walker, & Morwinski, 2013), and the third person effect (Douglas & Sutton, 2004 & 2008).

the majority of articles [they looked at] accepted the framing of a pause and sought to explain its cause. Furthermore, [these] authors often framed the article by juxtaposing the continuing increase of atmospheric CO2 levels with the presumed lack of warming on a decadal scale as though this presented a notable scientific problem at odds with expectations from greenhouse theory (Lewandowsky et al. 2016, 729). The above analysis from Lewandowsky, Risbey, and Oreskes 7

2

For example, see Kevin Elliot's A Tapestry of Values (2017), chapter 6.

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

Fig. 1. From Medhuag et al. (2017, 42; caption to Fig. 1b): “Peer-reviewed studies published annually (histogram) by the end of 2016 that contributed to the understanding of the hiatus (178 papers in total excluding ‘news and views’ and commentaries) and monthly output from ‘Google trends’ for the search criteria “global warming pause” [red] and “global warming hiatus” [blue], normalized to the maximum number of monthly searches for “global warming pause” ”. Papers continued to be published on the “hiatus” or ‘pause’ even after Medhaug et al.’s search, as described in Lewandowsky et al. (2018) who “searched the literature for articles (published through 2016) that referred to a “pause” or “hiatus” in GMST in the title or abstract. The search was completed in December 2017 and yielded 225 peerreviewed articles (see Risbey et al., 2018 for a complete list)” (3).

demonstrates that climate scientists used the “hiatus” or “pause” terms in a way that suggested acceptance—if only temporary—of the legitimacy of the ‘hiatus’. Part of the reason for this acceptance is likely due to both seepage and the susceptibility of climate scientists to certain cognitive biases (see footnote 6). However, it should be kept in mind that many climate scientists may have also been using the “hiatus” framing so as to present a problem for which their research-interest—their hobby-horse, so to speak—may be especially equipped to handle. Some examples will help demonstrate this hobby-horse explanation. First off, Lewandowsky et al. ((2015) and Lewandowsky et al. (2016)) see England et al. (2014) as adopting the “hiatus” framing. Indeed England et al. (2014) note that “Despite ongoing increases in atmospheric greenhouse gases, the Earth's global average surface air temperature has remained more or less steady since 2001”, (222) which suggests that the “hiatus” is a serious problem, worthy of consideration. However, this paper could as easily be seen as presenting such a frame for rhetorical purposes, to set the stage for their topic of investigation: how Pacific trade winds relate to temperature change in the context of inter-decadal Pacific oscillation. Pacific trade winds, or the inter-decadal Pacific Oscillation, can be understood as the hobby-horse or research interest of these scientists. On this view, the adoption of the “hiatus” framing can be understood as a strategy for making their paper seem relevant to broader public concerns, namely the alleged “hiatus”. Since there is a large body of literature supporting Lewandowsky et al.’s claim that scientists adopted the “hiatus” framing due to contrarian discourse, which supports Lewandowsky et al.’s seepage analysis, I suggest that this hobby-horse explanation is complementary to Lewandowsky et al.’s seepage analysis and is a significant part of the story that may have exacerbated the influence of contrarian discourse on the climate science community.8 The same hobby-horse explanation seems to work in many of the other papers Lewandowsky et al. (2016) assess as well. For example, Trenberth and Fasullo (2013) use the supposed “hiatus” as an occasion to discuss the global energy budget (and they also mention trade winds and the Pacific decadal oscillation) and Kosaka and Xie (2013) seem to only adopt the “hiatus” in order to reconcile model-observation discrepancies by “accounting for recent cooling in the eastern equatorial Pacific” (403 (abstract)). A more general example also demonstrates the hobby-horse explanation: a

climate scientist who focuses on ocean heat may be compelled to see how ocean heat uptake contributed to the apparent “hiatus” because this scientist or group of scientists work on ocean heat. That's their hobby-horse—it's what they do—so of course they would pursue research aimed at tying the supposed “hiatus” to ocean heat uptake. The point here is that many climate scientists may have adopted the “hiatus” framing partially for rhetorical purposes, to explore an apparent problem—i.e. the “hiatus”, a problem that was a “hot topic” in the media and was thus worth discussing and investigating—by pursuing their hobby-horse. Thus, this hobby-horse explanation may account for some of the disproportionate focus on the “hiatus” by the climate science community. Of course, this hobby-horse explanation is only part of the story. Some of the “hiatus” publications included in Fig. 1 above certainly adopted the “hiatus” framing for extra-rhetorical (that is, genuine) purposes. The literature cited in footnote 8 reinforces this point. However, the hobby-horse explanation seems to highlight a previously overlooked aspect of seepage, namely that the opportunity for hobby-horse riding may have exacerbated the impact of contrarian discourse on the climate science community. In addition to being susceptible to certain cognitive biases and political pressures, climate scientists may have also seen the “hiatus” as an occasion to pursue and publish on their research interest—a legitimate scientific interest—and thus they adopted the “hiatus” language and framing, partially for rhetorical purposes, in doing so. 3.2. ‘Climategate’ and the role of scientists as communicators While the hobby-horse explanation calls into question the sense in which climate scientists adopted the “hiatus” framing, the disproportionate focus on the “hiatus” can be explained in part by reaction to “Climategate”, and the role of climate scientists as communicators. As Medhaug et al. (2017) note, ‘Climategate’ likely contributed to the large amount of attention spent on the ‘hiatus’.9 More generally, they note that: “[W]ith past events in mind, scientists felt like they needed to respond to the public's distrust. Consequently, several of the early scientific studies related to the hiatus were motivated by the

9

‘Climategate’ was a 2009 controversy involving stolen emails from the Climate Research Unit and the University of East Anglia that was claimed by contrarians to show a climate science conspiracy. The climate scientists involved were exonerated (exonerated multiple times, in multiple investigations) of any wrongdoing/misconduct.

8

E.g. Brysse et al. (2013); Freudenburg and Muselli (2010); Biddle and Leuschner (2015); Leuschner (2016); and Mann (2015). Thanks to an anonymous reviewer for suggesting these sources and helping me think through the role of the hobby-horse explanation. 3

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

increased focus in the media and blogosphere, and focused on natural variations in the climate system as an explanation. Other reasons might also have had important roles in bringing the hiatus into public focus, and in making it not only a scientifically driven research topic, but also a publicly driven one, such as the argued lack of clear communication from scientists and the misleading wording of a ‘pause’ in global warming” (2017, 41).

4. Objectivity and the “hiatus” The key question to be addressed is thus, how has the objectivity of climate science been threatened and/or damaged due to the seepage of contrarian discourse? Posing this question emphasizes the point that while Lewandowsky et al. focus on the causes and existence of seepage, one of the aims in this paper is to elucidate the possible effects of seepage.12 I contend that the effects of seepage on climate science will become clearer by thinking in terms of objectivity. I've chosen Lloyd and Schweizer's seven meanings of objectivity because they were shown to be effective for comparing methodologies for determining future socioeconomic scenarios (Lloyd and Schweizer 2014).13 These different meanings of objectivity reflect different ways in which the term is used, and I apply these different meanings of objectivity to evaluate the different effects of seepage on the climate science community. Lloyd and Schweizer's seven meanings of objectivity are given in Table 1.

More generally, the attention paid to the “hiatus” by the IPCC is understandable in light of the IPCC's aim to provide policy-relevant assessments and reports to relevant government agencies and their aim of “Engaging and building relationships with the media” (IPCC Communications Strategy, 2012).10 Climate scientists are not only scientists but are also science communicators. The very existence of IPCC reports, especially the summary for policymakers, and the stated aim of the IPCC quoted above are evidence of this. Thus, it is reasonable that

Table 1 Lloyd and Schweizer's 7 Meanings of Objectivity. Each box represents a different sense of objectivity. So, for example, the top left box describes objectivity in the sense of being publicly accessible. If a piece of evidence or some experimental results are widely accessible (in a clear format) to the public, then it may be said to be objective in this sense. Methodological

Social

Metaphysical

1. Public: having to do with public accessibility, observability, and inter-subjectivity

4. Interactive: The notion of agreement achieved by intense debate/discussion among peers of the scientific community. Emphasis is on the degree to which both its procedures and its results are responsive to criticism (Longino, 1990). 5. Procedural: Achieved when a process is set up such that its outcome does not depend on who is performing that process (Douglas, 2007, p. 134; drawn from; Megill & Megill, 1994; Porter, 1992).

6. Independently existing: Independently existing from humans; phenomena that are not dependent on our perception for their existence.

2. Unbiased (understood statistically): A person may be unbiased even if they're making random mistakes or aren't detached. A person may still be biased even if they're committed to remaining unbiased: “biases such as availability and overconfidence can still affect the outcomes of a study or experiment, because the mental processes are operating below the conscious level … Such cognitive biases are systematic, particular, and undesired” (Lloyd and Schweizer 2014, 2068). 3. Detached: disinterested, independent from will or wishes. To be detached, values shouldn't supplant evidence

7.“Really real”: Distinct from independently existing; for example, colors and dreams

climate scientists may have wanted to alleviate the worries and doubts expressed in the public and in the media by explaining—or explaining away—the alleged “hiatus”. In sum, in response to contrarian discourse and its influence on the general public, policymakers, and the media, climate scientists may have felt a responsibility to expend extra effort in the form of more publications and research and an inclusion of the “hiatus” in the IPCC's Fifth Assessment Report. This extra effort not only led to greater scientific knowledge (see next section) but also provided an overwhelming body of research refuting basic contrarian claims concerning the “hiatus”, thus reassuring the public.11 With a fuller understanding of seepage and the global warming “hiatus” now in place, in the next section I will provide a more holistic and systematic account of the effects of seepage by examining the objectivity of the climate science community.

First note that this list may not be exhaustive—Douglas (2004) for example discusses concordance of results and convergence of different approaches reaching the same result, both of which overlap to some extent with “procedural objectivity” in Table 1 above. Note also that these meanings are not completely distinct from each other. For example, if a methodology is more objective in each of the methodological and social senses, it may be more likely that its results are informative of independently existing phenomena. Additionally, if interactive objectivity in a community is achieved to a high degree then this may counteract a lack of detachment—scientists who previously supplanted evidence with their values would respond to critiques and discussion within their community if interactive objectivity is met to a high degree. I'll first discuss the methodological meanings from Table 1, proceeding from #1 to #3. Then I'll turn to the social senses, #4 and

10 Admittedly, the importance of the IPCC as an international body that issues regular reports suggest that the events leading up to the “hiatus” being included in the IPCC's Fifth Assessment Report are worthy of an additional investigation. As will be noted in section 4, this seems like a failure to meet the criteria of interactive objectivity, however the exact treatment of the “hiatus” term, whether the “hiatus” framing was used, and the politics surrounding the decision to include the “hiatus” in such an important report, all suggest that there's more to be said about this. 11 For discussion of climate scientists communicating the “hiatus” issue to the public at an IPCC press conference, and some potential problems, see Hollin and Pearce (2015) and a critical response by Jacobs et al. (2015).

12 While Lewandowsky et al. do not set out to elaborate the full effects of seepage, they do mention one effect of seepage: Victor and Kennel (2014), who, largely because of the alleged ‘hiatus,’ suggest abandoning the 2 °C warming goal and focus instead “on a set of more diverse measures of climate change and its impacts” (Lewandowsky et al., 2015, p. 8). 13 Cross Impact Balance (CIB) method was shown to be a better method than Intuitive Logics for projecting future socioeconomic scenarios. That is, CIB was shown to be more objective in multiple senses of the term. The superior objectivity of CIB was used as grounds for recommending its use for Integrated Assessment (IA) models, which are models used by the IPCC. Importantly, “Since the First Assessment Report of the IPCC, IA model results have served as the working assumptions for general circulation models (Hibbard et al., 2007)” (Lloyd and Schweizer 2014, 2051).

4

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

#5, and then I'll discuss the metaphysical meanings, #6 and #7. Seepage does seem to have negatively impacted the public accessibility of climate science (objectivity #1) because it led to a disproportionate focus on the “hiatus”, thereby contributing to public misinformation concerning the legitimacy of the “hiatus”. In an attempt to alleviate public worries about the alleged contrarian claim that climate models were fundamentally mistaken, climate scientists likely increased levels of public confusion concerning the “hiatus”.14 That is, being told that there is no “pause” or “hiatus” and then seeing 200 papers published on this non-phenomenon would reasonably cause some confusion for even scientifically literate segments of the population, even though none of the information concerning GMST trends was, strictly speaking, inaccessible to the public. Moving to objectivity #2, unbiased, the large number of publications that focus on the “hiatus” (see Fig. 1) is evidence of bias. There was clearly a biased focus on the “hiatus” time period as a topic to investigate and publish on—and this is one of the most evident impacts of seepage. It is indisputable that climate scientists paid more attention to the 1998–2012 time period than to any other time periods of comparable length, even periods of accelerated warming, all of which constitutes bias.15 The type of bias playing out here is not a cognitive bias, such as the ones discussed by Lewandowsky et al., it is a statistical sense of bias. Even though Lloyd and Schweizer (2014) give several examples of cognitive biases and their outcomes, their description of “unbiased” needn't be restricted to these examples. In the “hiatus” case study the evident bias can be understood as follows: If all climate science publications were collected and grouped together according to the topics they addressed, including the length of time periods or other statistical fluctuations they discussed, and the results of this grouping were graphed or otherwise evaluated, there would be a clear bias towards those papers focusing on the “hiatus” time period. In the context of what this means for how seepage impacted climate science, a relevant question here might be: if climate scientists had not published so heavily on the “hiatus”, what else would they have been researching? Also, what other knowledge might have been gained in exchange for the knowledge that was gained in these “hiatus” publications? Of course, this is an impossible-to-answer counterfactual question, but it highlights the concern that perhaps climate scientists might have spent their time and energy on more fruitful research had there been no contrarian influence. Related to this issue is the hobby-horse explanation, described in section 3.1. It seems that since the “hiatus” was a topic of concern broadly—in the media and in the scientific community—it may have incentivized hobby-horse riding. The existence of more and more peer-reviewed papers on the “hiatus” creates the impression that journals are ready and willing to publish research on this topic. This may reflect an interesting feature of scientific research: what gets published has an impact on what is researched. In some cases, this can promote hobby-horse riding. Climate scientists were not fully detached (objectivity #3) in this “hiatus” episode, but this issue requires careful treatment. Climate scientists failed to be fully detached in the sense that social and ethical values associated with reassuring the public played a role in which topics were researched, how the research was communicated, and ultimately why the “hiatus” was included in the IPCC's Fifth Assessment Report (although this issue may not be so simple, see footnote 10 above). Values associated with individual research interests—which are legitimate, evidence-based scientific interests—also seem to have played a role. Many publications focused on the time period, and the “hiatus” and “pause” terms were heavily used, but the publications that focus on the “hiatus” do not supplant evidence with values—assuming the reliability of the peer-review process—and in this respect climate scientists remained detached in performing and evaluating their 14 15

scientific work. Thus, the climate science community does not seem to have been compromised by contrarian influence when it comes to the reliability of their published conclusions. The use of the “hiatus” term and the adoption of the “hiatus” framing (based on Lewandowsky et al., 2016) are evidence that scientists were not detached when deciding how to frame a research article or decide a research question, but this is importantly distinct from supplanting evidence with values, which scientists did not apparently do. The inclusion of the “hiatus” in the IPCC's Fifth Assessment Report (2013) is evidence that interactive objectivity (objectivity #4) of the IPCC has also been damaged by seepage. Given the IPCC's goal to provide policy-relevant information, there will always be extra-scientific concerns that influence what is included in a given report. However, the focus on the “hiatus” time period and use of the term were both criticized by the scientific community, and yet the “hiatus” was still included in the report to a significant degree—an entire “box” of the Working Group 1 Report!16 In this case it seems that contrarian voices may have influenced decisions that should've been made exclusively by the scientific community. At the very least, criticisms about discussing the “hiatus” in the report were not given sufficient attention, leading to an entire “box” discussing the “hiatus” in the report, even though the “hiatus” was a demonstrably statistically insignificant fluctuation in GMST. Whether the report should have mentioned the “hiatus” at all is a separate question from whether or not the scientific community was appropriately responsive to criticism regarding the inclusion of the “hiatus” in the report. For example, the “hiatus” could have been included in the report with a disclaimer explicitly pointing out its statistical insignificance and the origin of the “hiatus” as a piece of contrarian rhetoric based on contrarians cherry-picking a time period that begins with an anomalously warm GMST. However, the scientific community seems to have taken Lewandowsky, Oreskes, Risbey, Newell, and Smithson (2015)'s claims about seepage seriously, because papers published after the seepage paper which mention the “hiatus” seem to do so more carefully. For example, in “Making sense of the early 2000s warming slowdown” (2016), Fyfe et al. explicitly agree with Lewandowsky et al. that the “hiatus” framing is problematic. Moreover, Fyfe et al. (2016) agree that the short-term variations during the “hiatus” time period are statistically insignificant: “Our goal here is to move beyond purely statistical aspects of the slowdown, and to focus instead on improving process understanding and assessing whether the observed trends are consistent with our expectations based on climate models” (2016, 226). The process understanding they focus on has to do with the inter-decadal Pacific oscillation (IPO), which was a topic also investigated by England et al. (2014), mentioned in section 3.1 above.17 Medhaug et al. (2017) also cite Lewandowsky et al. (2015) before carefully outlining several candidate definitions of the “hiatus”, surveying the scientific literature in which the “hiatus” is discussed, and using new datasets to ultimately conclude the following: “Most discrepancies between models and observations can therefore be explained by the state of the natural variability, incomplete or biased forcings, and observational limitations; a complete explanation requires a combination of all of these” (45). What these examples demonstrate is that climate scientists took the criticism of Lewandowsky et al. (2015) seriously, which is evidence that the climate science community is interactively objective to an extent, despite the problems with the IPCC's Fifth Assessment Report, mentioned above.18 In contrast, arguably one of the mistakes made by

16 Box 9.2, pages 769–772 of the Working Group 1 Report, “The Physical Science Basis”. 17 In addition to process knowledge, knowledge about observational datasets was also gained (for example, see Karl, Anthony, Huang, LawrimoreMcMahonMennePetersonVose, & Zhang, 2015). 18 Rajaratnam, Romano, Tsiang, Noah, and Diffenbaugh (2015) and Yan et al. (2016) also explicitly reject the “hiatus” framing. However, it may just be that

Thanks to two anonymous reviewers for this point. For more on this see Lewandowsky et al. (2016). 5

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

climate scientists was taking the contrarian critique—the idea that there was a “hiatus” in warming—seriously. This highlights an important point made by Helen Longino that, while legitimate criticisms can come from outside a scientific community (or from a knowledge community more broadly, which can be thought of as including nonscientific members), the community isn't responsible for responding to illegitimate critiques or bad faith skepticism. In Longino's words, “reiterating the same old complaint no matter what response is offered eventually disqualifies one as a member of a discursive community of equals” (2002, 133). In the “hiatus” case study, critiques such as Lewandowsky et al.’s are legitimate, whereas contrarian sources of criticism and critique are clearly illegitimate. Procedural objectivity (objectivity #5) has also been damaged by seepage, and the different ways in which it has been damaged shows that there is an ambiguity to procedural objectivity when applying it to scientific communities. This is because there may be both scientific and non-scientific procedures involved. For example, the IPCC is an international body of scientists and thus it is a scientific community, but its decision to include the “hiatus” in the Fifth Assessment Report is not the result of a purely scientific procedure, it is a procedure concerned primarily with communication to the public and to policymakers. However, this procedure likely does depend on who is performing it, because at least some climate scientists disagreed with the decision to include the “hiatus” in the report. It is unsurprising that social procedures such as these are not completely procedurally objective, but we could also consider the counterfactual scenario in which there was no contrarian discourse or “hiatus” framing, and in such a scenario the “hiatus” would definitely not have been in the report. Thus, it seems that seepage has highlighted a sense in which climate science is not procedurally objective, but this lack of objectivity does not seem to be a direct result of seepage. In contrast, there is a straightforward way in which seepage does seem to have damaged the procedural objectivity of climate science research, because some papers were careful to avoid (or were critical of) the “hiatus” framing, whereas others were not, and yet all of these papers went through the peer-review process. One of the functions of peer-review procedures in scientific journals is to provide a gatekeeping function, and this gate-keeping function includes preventing or minimizing problematic frames, such as the “hiatus” frame. The peerreview process did not achieve this goal in the “hiatus” case, which is evident from the “considerable heterogeneity among authors of the extent to which they have adopted the “pause” frame” (Lewandowsky et al., 2015, p. 6).19 This heterogeneity among authors reflects at least some heterogeneity among reviewers thus highlighting the sense in which peer-review is not procedurally objective. It may be unreasonable to expect peer-review procedures to be completely objective—to not depend at all on who is performing the procedure—but it does not seem unreasonable to expect peer-review to avoid systematic problems like failing to screen out certain misleading terms and frames that originated from contrarian discourse. As such, seepage has significantly damaged the procedural objectivity of the climate science community with this episode. Additionally, the case of the “hiatus” presents strong evidence that evaluating rhetorical set-ups and frames may matter just as much as evaluating the data, methods, and conclusions, because

peer-review procedures that fail to screen them out can have negative consequences (i.e. see discussion of public accessibility above). The increased presence of the “hiatus” framing in the published literature may have also contributed to hobby-horse riding, as mentioned above, thereby exacerbating the amount of attention paid to the “hiatus”. Lastly, the independent existence (objectivity #6) and the reality (objectivity #7) of climate process knowledge gained from research that focused on the “hiatus” was not damaged by seepage. This is a point of defense worth stressing: that the conclusions of climate scientists in their published research are still reliable even though they have been influenced by contrarian discourse. Much of this process knowledge reconciles the differences between observational data and climate models for 1998–2012 and it was arrived at objectively in the sense that evidence, rather than the wishes/wills of climate scientists, support the conclusions drawn. Moreover, the evaluation of this evidence is both procedurally and interactively objective because these publications were all peer-reviewed (although the point about gatekeeping mentioned above should not be forgotten). The climate processes may thus be said to be objective in the sense of independently existing because we have good reason to think that the conclusions drawn tell us about independently existing climate processes and are not a product of wishful thinking, statistical bias, or procedures whose results vary based on the subjective/character features of individual scientists, or the extent to which contrarian discourse has seeped into to those scientists' psyche. Whether these climate processes are also “really real” (objectivity #7) is a question that can be pursued by those who are more metaphysically inclined, but for now it suffices to say that the independent existence of climate processes such as the IPO or equilibrium climate sensitivity (how the atmosphere's temperature changes as a function of CO2 doubling), for example, seem to be real if they are independently existing.20 To have established the independent existence of climate processes relevant both to short-term temperature fluctuations and for future climate projections sufficiently captures an important point that climate science has not been compromised in a way that invalidates the conclusions its scientists have drawn. Of course, this should be understood in conjunction with the fact that seepage has occurred. In sum, the above analysis highlights the impact of contrarian discourse, how climate scientists have responded to the public's worries, how the “hiatus” was used as an opportunity to investigate and publish on topics of legitimate scientific interest, and how the conclusions drawn by climate scientists in their published work are reliable despite the reality of seepage. 5. Conclusion The evidence surveyed in this paper suggests that, while contrarian discourse has seeped into the climate science community, the knowledge of climate processes gained from focusing on the “hiatus” corresponds to independently existing processes and is therefore objective in that sense. The public, unbiased, interactive, and procedural senses of objectivity were all damaged to varying degrees, but it is crucial to emphasize that climate scientists did not engage in wishful thinking when evaluating evidence in their peer-reviewed research even though the peer-reviewed process failed to screen out problematic terms/ frames, and even though hobby-horse riding may have contributed to the disproportionate amount of research on the “hiatus”. In light of this, it is crucial to emphasize the climate science community's responsiveness to criticism that was discussed above, and their responsiveness to

(footnote continued) these authors became more confident in criticizing the “hiatus” framing because by 2015 it was clear that the “hiatus” had stopped and that it remained insignificant. Thanks to an anonymous reviewer for pointing this out. 19 Seneviratne, Donat, Mueller, and Alexander (2014), for example, were critical of the “hiatus”, or “pause”. Arguably, if these authors had been referees on one of the papers that used the “hiatus” framing less responsibly, they may have been critical of such use and sent such a paper back to the authors to change the frame or rhetorical setup. While this example is speculative it highlights the sense in which the procedural objectivity of the peer-review process was undermined in the “hiatus” case.

20

There is a sense in which the climate process knowledge we've gained from studying the “hiatus” corresponds to really real but not independently existing processes since without humans there would be no climate change to begin with. See Lloyd and Schweizer 2014, pages 2079–2083 for a discussion of the relationship between these senses of objectivity. 6

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin

criticism moving forward. For example, an important pair of papers were recently published (Lewandowsky et al., 2018; Risbey et al., 2018) that historically situate the focus on the “hiatus” time period in light of knowledge scientists had at particular times (which observational datasets were available when specific claims about the “hiatus” were made) and conclude that the “hiatus” was only a statistically legitimate phenomenon from 2011-13 and only if a selection bias was ignored.21 These papers provide further evidence of seepage and effectively criticize decisions made by climate scientists who decided to investigate short term trends that were statistically insignificant without paying sufficient attention to the frames they implicitly adopted. The responses of the scientific community to this new criticism will serve as an additional measure of their objectivity in several senses, including interactive, detached, procedural, and perhaps others. Thus, a full understanding of seepage's impact on the climate science community will require an ongoing consideration of the climate science community's objectivity, multiply understood.

Douglas, Heather (2007). “Rejecting the ideal of value-free science.” in value-free science?. In H. Kincaid, John Dupre, & Alison Wylie (Eds.). 120–39. Oxford University Press. Douglas, Karen M., & Sutton, Robbie M. (2004). Right about others, wrong about ourselves? Actual and perceived self-other differences in resistance to persuasion. British Journal of Social Psychology, 43(4), 585–603. https://doi.org/10.1348/ 0144666042565416. Douglas, Karen M., & Sutton, Robbie M. (2008). The hidden impact of conspiracy theories: Perceived and actual influence of theories surrounding the death of princess diana. The Journal of Social Psychology, 148(2), 210–221. https://doi.org/10.3200/ SOCP.148.2.210-222. Elliot, Kevin (2017). A Tapestry of values: An introduction to values in science. Oxford University Press. England, Matthew H., McGregor, Shayne, Paul, Spence, Meehl, Gerald A., Timmermann, Axel, Cai, Wenju, et al. (2014). Recent intensification of wind-driven circulation in the pacific and the ongoing warming hiatus. Nature Climate Change, 4(3), 222–227. https://doi.org/10.1038/nclimate2106. Freudenburg, William R., & Muselli, Violetta (2010). Global warming estimates, media expectations, and the asymmetry of scientific challenge. Global Environmental Change, Governance, Complexity and Resilience, 20(3), 483–491. https://doi.org/10.1016/j. gloenvcha.2010.04.003. Fyfe, John C., Meehl, Gerald A., England, Matthew H., Mann, Michael E., Santer, Benjamin D., Flato, Gregory M., et al. (2016). Making sense of the early-2000s warming slowdown.” Comments and Opinion. Nature Climate Change. February 24, 2016 https://doi.org/10.1038/nclimate2938. Hibbard, K. A., Meehl, G. A., Cox, P. M., & Friedlingstein, P. (2007). A strategy for climate change stabilization experiments. Eos, 88(2007), 217–221. Hollin, G. J. S., & Pearce, W. (2015). Tension between scientific certainty and meaning complicates communication of IPCC reports. Nature Climate Change, 5(8), 753–756. https://doi.org/10.1038/nclimate2672. IPCC. (2013). Climate change 2013: The physical science Basis. Contribution of working group I to the Fifth assessment report of the intergovernmental Panel on climate changeCambridge, United Kingdom and New York, NY, USA: Cambridge University Press. https://doi.org/10.1017/CBO9781107415324. Jacobs, Peter, Hunter, Cutting, Lewandowsky, Stephan, O'Brien, Miriam, Rice, Ken, & Verheggen, Bart (2015). Clarity of meaning in IPCC press conference. Nature Climate Change, 5(11), 961–962. https://doi.org/10.1038/nclimate2845. Karl, Thomas R., Anthony, Arguez, Huang, Boyin, Lawrimore, Jay H., McMahon, James R., Menne, Matthew J., et al. (2015). Possible artifacts of data biases in the recent global surface warming hiatus. Science. June, aaa5632 https://doi.org/10.1126/ science.aaa5632. Kosaka, Yu, & Xie, Shang-Ping (2013). Recent global-warming hiatus tied to equatorial pacific surface cooling. Nature, 501(7467), 403–407. https://doi.org/10.1038/ nature12534. Leuschner, Anna (2018). Is it appropriate to ‘target’ inappropriate dissent? On the normative consequences of climate skepticism. Synthese, 195(3), 1255–1271. https:// doi.org/10.1007/s11229-016-1267-x. Leviston, Z., Walker, I., & Morwinski, S. (2013). Your opinion on climate change might not Be as common as you think. Nature Climate Change, 3(4), 334–337. https://doi. org/10.1038/nclimate1743. Lewandowsky, Stephan, Kevin, Cowtan, Risbey, James S., Mann, Michael E., Steinman, Byron A., Oreskes, Naomi, et al. (2018). The ‘pause’ in global warming in historical context: (II). Comparing models to observations. Environmental Research Letters, 13(12), 123007. https://doi.org/10.1088/1748-9326/aaf372. Lewandowsky, Stephan, Oreskes, Naomi, Risbey, James S., Newell, Ben R., & Michael Smithson (2015). Seepage: Climate change denial and its effect on the scientific community. Global Environmental Change, 33(July), 1–13. https://doi.org/10.1016/j. gloenvcha.2015.02.013. Lewandowsky, Stephan, Risbey, James S., & Oreskes, Naomi (2016). “The ‘pause’ in global warming: Turning a routine fluctuation into a problem for science. Bulletin of the American Meteorological Society, 97(5), 723–733. https://doi.org/10.1175/BAMSD-14-00106.1. Lloyd, Elisabeth A. (1995). Objectivity and the double standard for feminist epistemologies. Synthese, 104(3), 351–381. https://doi.org/10.1007/BF01064505. Lloyd, Elisabeth A., & Schweizer, Vanessa J. (2014). Objectivity and a comparison of methodological scenario approaches for climate change research. Synthese, 191(10), 2049–2088. https://doi.org/10.1007/s11229-013-0353-6. Longino, Helen (1990). Science as social knowledge. Princeton University Press. Longino, Helen E. (2002). The fate of knowledge. Princeton University Press. Mann, Michael E. (2015). The serengeti strategy: How special interests try to intimidate scientists, and how best to fight back. Bulletin of the Atomic Scientists, 71(1), 33–45. https://doi.org/10.1177/0096340214563674. McCright, A. M. (2007). In S. C. Moser, & L. Dilling (Eds.). “Dealing with climate change contrarians” in Creating a Climate for change: Communicating climate Change and facilitating social change (pp. 200–212). New York: Cambridge University Press. Medhaug, Iselin, Stolpe, Martin B., Fischer, Erich M., & Knutti, Reto (2017). Reconciling controversies about the ‘global warming hiatus.’. Nature, 545(7652), 41–47. https:// doi.org/10.1038/nature22315. Megill, A., & Megill, A. (1994). Introduction: Four senses of objectivity. Rethinking objectivity: Vols. 1–20Duke University Press. Mooney, Chris (2013). Who created the global warming ‘pause’? Mother Jones. October 7, 2013 https://www.motherjones.com/environment/2013/10/global-warming-pauseipcc/. O'Neill, Saffron J., & Boykoff, Max (2010). Climate denier, skeptic, or contrarian? Proceedings of the National Academy of Sciences, 107(39), E151–E151 https://doi.org/ 10.1073/pnas.1010507107. Porter, Theodore M. (1992). Quantification and the accounting ideal in science. Social

Declaration of interest statement This author declares no competing financial interests. Acknowledgements Thanks to Lisa Lloyd for helpful comments and suggestions throughout the various phases of this paper and for helping me make some important connections in the climate science community. Additional thanks to Stephan Lewandowsky for answering my questions about his research and for giving me feedback on an earlier version of this project. Thanks also to the members of Lloyd lab and various passersby at the 2018 PSA and 2018 AGU conference poster sessions, and thank you to two anonymous reviewers who provided excellent feedback on an earlier draft. A special thanks to Dan Li for reading so many drafts of this paper and for encouragement throughout. References IPCC communications strategy: Adopted by the Panel at the thirty-fifth session (Geneva, 6–9 June 2012), Amended at the forty-fourth session (Bangkok, 17-20 October 2016). n.d https://www.ipcc.ch/meeting_documentation/pdf/Communication/IPCC_ Communications_Strategy.pdf. Nature Geoscience157. Nature Climate Change149. Biddle, Justin B., & Leuschner, Anna (2015). Climate skepticism and the manufacture of doubt: Can dissent in science Be epistemically detrimental? European Journal for Philosophy of Science, 5(3), 261–278. https://doi.org/10.1007/s13194-014-0101-x. Brysse, Keynyn, Oreskes, Naomi, O'Reilly, Jessica, & Oppenheimer, Michael (2013). Climate change prediction: Erring on the side of least drama? Global Environmental Change, 23(1), 327–337. https://doi.org/10.1016/j.gloenvcha.2012.10.008. Dirikx, Astrid, & Gelders, Dave (2010). To frame is to explain: A deductive frame-analysis of Dutch and French climate change coverage during the annual UN conferences of the parties. Public Understanding of Science, 19(6), 732–742. https://doi.org/10.1177/ 0963662509352044. Douglas, Heather (2004). The irreducible complexity of objectivity. Synthese, 138(3), 453–473. https://doi.org/10.1023/B:SYNT.0000016451.18182.91.

21 “However, therein lies a problem: if a period is chosen (from many possible such time intervals) because of its unusually low trend, this has implications for the interpretation of conventional significance levels (i.e. p-values) of the trend (Rahmstorf, Foster, & Cahill, 2017). Selection of observations based on the same data that is then being statistically tested inflates the actual p-value, thereby giving rise to a larger proportion of statistical Type I errors than the researcher is led to expect (Wagenmakers, 2007). Very few articles on the ‘pause’ account for or even mention this effect, yet it has profound implications for the interpretation of the statistical results” (Lewandowsky et al., 2018, p. 8). In the same paper, from the conclusion: “During the period 2011–2013, the impression of a divergence could appear to be statistically significant, but only if the selection-bias issue was ignored. We have shown that ignoring of the selection bias issue can drastically inflate Type-I error rates, which renders the inferences unreliable and in this case erroneous” (23).

7

Studies in History and Philosophy of Science xxx (xxxx) xxx–xxx

R. O'Loughlin Studies of Science, 22(4), 633–651. https://doi.org/10.1177/030631292022004004. Rahmstorf, Stefan, Foster, Grant, & Cahill, Niamh (2017). Global temperature evolution: Recent trends and some pitfalls. Environmental Research Letters, 12(5), 054001 https://doi.org/10.1088/1748-9326/aa6825. Rajaratnam, Bala, Romano, Joseph, Tsiang, Michael, Noah, S., & Diffenbaugh (2015). Debunking the climate hiatus. Climatic Change, 133(2), 129–140. https://doi.org/10. 1007/s10584-015-1495-y. Risbey, James S., Lewandowsky, Stephan, Kevin, Cowtan, Oreskes, Naomi, Rahmstorf, Stefan, Jokimäki, Ari, et al. (2018). A fluctuation in surface temperature in historical context: Reassessment and retrospective on the evidence. Environmental Research Letters, 13(12), 123008. https://doi.org/10.1088/1748-9326/aaf342. Seneviratne, Sonia I., Donat, Markus G., Mueller, Brigitte, & Alexander, Lisa V. (2014). No

pause in the increase of hot temperature extremes. Nature Climate Change, 4, 161–163. February https://doi.org/10.1038/nclimate2145. Trenberth, Kevin E., & Fasullo, John T. (2013). An apparent hiatus in global warming? Earth’s Future, 1(1), 19–32. https://doi.org/10.1002/2013EF000165. Victor, David G., & Kennel, Charles F. (2014). Climate policy: Ditch the 2 °C warming goal. Nature News, 514(7520), 30. https://doi.org/10.1038/514030a. Wagenmakers, Eric-Jan (2007). A practical solution to the pervasive problems of p values. Psychonomic Bulletin & Review, 14(5), 779–804. https://doi.org/10.3758/ BF03194105. Yan, Xiao-Hai, Boyer, Tim, Trenberth, Kevin, Karl, Thomas R., Xie, Shang-Ping, Nieves, Veronica, et al. (2016). The global warming hiatus: Slowdown or redistribution? Earth’s Future, 4(11), 472–482. https://doi.org/10.1002/2016EF000417.

8