I need to write 5 bibliographies for an essay I’m doing discussing how we shouldn’t use animals for cosmetic testing. I included the 5 sources and 2 files that explain how it should be done and an exa

I need to write 5 bibliographies for an essay I’m doing discussing how we shouldn’t use animals for cosmetic testing. I included the 5 sources and 2 files that explain how it should be done and an exa
The Cosmetics Europe strategy for animal-free genotoxicity testing: Project status up-date S. Pfuhler a,⇑, R. Fautz b, G. Ouedraogo c, A. Latil d, J. Kenny e, C. Moore f, W. Diembeck g, N.J. Hewitt h, K. Reisinger i, J. Barroso j aProcter & Gamble Co., 1 Procter and Gamble Plz, Cincinnati, OH, USAbKao Germany GmbH, Pfungstädterstraße 92 – 100, D-64297 Darmstadt, GermanycL’Oreal Life Sciences Research, Aulnay sous Bois, FrancedPierre Fabre, 3 Rue des satellites, 31432 Toulouse, FranceeGSK, Park Road, Ware, SG12 ODP, UKfUnilever, Colworth House, Sharnbrook, Bedford, MK44 1LQ, UKgBeiersdorf AG, Unnastrasse 4, D-20253 Hamburg, GermanyhSWS,Wingertstrasse 25, D-64390 Erzhausen, GermanyiHenkel AG & Co., KGaA, Henkelstraße 67, D-40191 Duesseldorf, GermanyjCosmetics Europe, Avenue Herrmann Debroux 40, B-1160 Auderghem, Brussels, Belgium article info Article history: Received 11 October 2012 Accepted 18 June 2013 Available online 27 June 2013 Keywords: Genotoxicity Cosmetics Alternatives to animal testing 3D Skin Metabolism abstract The Cosmetics Europe (formerly COLIPA) Genotoxicity Task Force has driven and funded three projects to help address the high rate of misleading positives inin vitrogenotoxicity tests: The completed ‘‘False Positives’’ project optimized current mammalian cell assays and showed that the predictive capacity of thein vitromicronucleus assay was improved dramatically by selecting more rel- evant cells and more sensitive toxicity measures. The on-going ‘‘3D skin model’’ project has been developed and is now validating the use of human reconstructed skin (RS) models in combination with the micronucleus (MN) and Comet assays. These models better reflect the in use conditions of dermally applied products, such as cosmetics. Both assays have demonstrated good inter- and intra-laboratory reproducibility and are entering validation stages. The completed ‘‘Metabolism’’ project investigated enzyme capacities of human skin and RS models. The RS models were shown to have comparable metabolic capacity to native human skin, confirming their usefulness for testing of compounds with dermal exposure. The program has already helped to improve the initial test battery predictivity and the RS projects have provided sound support for their use as a follow-up test in the assessment of the genotoxic hazard of cos- metic ingredients in the absence ofin vivodata. 2013 Elsevier Ltd. All rights reserved. 1. Introduction The focus of many researchers has been to develop better in vitrotools to replace animal tests, especially in the light of reg- ulations such as the 7th amendment to the Cosmetics Directive (EU, 2003) and REACh (European Commission, 2006). The use of in vitromodels is especially relevant to the cosmetics industry which is banned from using animal tests for a number of end- points, including genotoxicity. Thus, positive outcomes in standard in vitroassays evaluating the genotoxic potential of chemicals can no longer be followed-up within vivoassays. If genotoxicity is as- sessed using onlyin vitroassays, this may well result the de-selec- tion of many safe new products since these assays have a high rate of positive results that do not correlate within vivogenotoxicity or carcinogenicity (Kirkland et al., 2005). This problem is recognized as a critical issue and has led to a number of working groups inves- tigating improved approaches for assessing genotoxicity (e.g., International Life Sciences Institute – Human and Environmental Sciences Institute’s (ILSI-HESI) committee on The Relevance and Follow-up of Positive Results inin vitroGenotoxicity Testing (IVGT); www.hesiglobal.org). As part of an international and mul- ti-laboratory collaboration, Cosmetics Europe (formerly ‘‘COLIPA’’) has funded and driven numerous projects aimed to address the lack of adequate alternatives to traditionalin vivotests and help validate successful models. 0887-2333/$ – see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tiv.2013.06.004 ⇑Corresponding author. Tel.: +1 513 319 7468. E-mail addresses:[email protected](S. Pfuhler),[email protected](R. Fautz), [email protected](G. Ouedraogo),[email protected](A. Latil), julia.d.ken[email protected](J. Kenny),[email protected](C. Moore),Walter. [email protected](W. Diembeck),[email protected](N.J. He- witt),kerstin.r[email protected](K. Reisinger),[email protected](J. Barroso). Toxicology in Vitro 28 (2014) 18–23 Contents lists available atSciVerse ScienceDirect Toxicology in Vitro journal homepage:www.elsevier.com/locate/toxinvit The Cosmetics Europe Genotoxicity Task Force was set up to coordinate and drive three main projects. The aim of the ‘‘False Positives’’ project was to optimize current mammalian cell assays by focusing on two aspects of the micronucleus (MN) test, namely the cell type employed and the method of cytotoxicity measure- ment. The ‘‘3D skin model’’ project aimed to develop and validate a new assay incorporating the use of human reconstructed skin (RS) models with the genotoxicity endpoints, MN and Comet as- says. Since the exposure of many compounds is dermal, especially for cosmetics, these models better reflect their in use conditions. In order to interpret outcomes from the 3D model assays, knowledge of the metabolic capacity of these RS models is an advantage, espe- cially in comparison with native human skin. Therefore, the ‘‘Metabolism’’ project investigated the enzyme capacities of human skin and compared them with that in RS models and with 2D monolayer cultures of skin cells. Here, we review the outcomes of the Genotoxicity Task Force projects, outline future research aims and apply the knowledge gained to a decision tree approach to assess the genotoxic potential of chemicals used in the cosmetics industry. 2. The ‘‘False Positives’’ project One of the drawbacks of the currentin vitroclastogenicity as- says is that they produce many positive results when compared with negative rodent carcinogenicity data. Indeed, an analysis of published data revealed the rate of misleading positives to be at least 80% when threein vitroassays were combined into a test bat- tery (Kirkland et al., 2005), as requested by several guidelines (such as ReachEuropean Commission, 2006and the SCCS notes of guid- anceSCCP, 2009). Therefore, existing and new assays need to show better specificity (i.e. correctly identify non-carcinogens) without compromising sensitivity (i.e. still detectingin vivogenotoxins and DNA-reactive carcinogens). The ‘‘False positive’’ project was set up to optimize thein vitroMN test such that the methodology described in the OECD guideline 487 (OECD, 2010) was retained. A European Center for the Validation of Alternative Methods (EC- VAM) workshop held in 2006 discussed ways to reduce the fre- quency of misleading positive results. Several suggestions for possible improvements/modifications to existing tests were identi- fied, including the cell lines used and the method of cytotoxicity measurement (Kirkland et al., 2007). Following from these recom- mendations we started a program which was executed at Covance Laboratories, UK, where we investigated how the predictive capacity of thein vitroMN test is impacted by the choice of cell lines (Fowler et al., 2012a) and toxicity parameters (Fowler et al., 2012b). Others have compared different cell types (Sofuni et al., 1990; Hilliard et al., 2007; Erexson et al., 2001) but this is the first comparison that includes six cell types within the same project and using consistent, GLP-like experimental conditions including the same batches of chemicals, media and formulations (Fowler et al., 2012a). These investigations revealed that certain cell types were more prone to misleading positive responses than others, particularly rodent cell lines (Fig. 1). The p53 compromised rodent cell lines, CHL, CHO and V79, demonstrated poorer specificity than the p53 functional human cell types, TK6, HepG2 and human peripheral blood lympho- cytes (HuLy) in response to compounds accepted as producing mis- leading positive results inin vitroclastogenicity assays (Kirkland et al., 2008). Cells of human origin may also be more favorable than rodent cells since these are more representative of human responses and may contain more human-specific metabolising enzymes and transporter proteins than rodent-derived cell types. Improving specificity cannot be at the expense of compromised sensitivity and thus it was necessary to make sure that the choice of cells did not lead to a decrease in sensitivity. Seventeen carcinogens that are thought to act via genotoxic mechanism (from the Group 1 chemicals fromKirkland et al., 2008) were re-tested in human lym- phocytes and TK6 cells. Overall the data show that out of a panel of 17 genotoxic chemicals, TK6 and HuLy detect the majority of them as positive (15 out of 17; 88% accuracy), confirming the high sensitivity of these cells (Fowler et al., 2013). Investigations of the cell type employed the replication index (RI) as a measure of cytotoxicity; however, it was considered that cytotoxicity may also be a parameter that could influence the out- come of the assay since this is used to select the top concentration tested (Kirkland et al., 2007). For example, one study showed that Relative Cell Counts (RCC) underestimated the toxicity of a number of direct and indirect genotoxins and thus selected higher concen- trations for subsequent MN analysis (Fellows et al., 2008). There- fore, additional measures of cytotoxicity of a number of misleading positives in CHO, CHL and TK6 cells were investigated in the ‘‘False Positives’’ project, namely Relative Population Dou- bling (RPD), Relative Increase in Cell Counts (RICC), RI and RCC. Re- sults revealed that estimation of toxicity based on relative proliferation increases (RPD and RICC) tended to select concentra- tions in the target toxicity range (50–60%) that give mainly nega- tive MN responses (Fig. 2). Conversely, measurements of RCC and RI selected concentrations in the target toxicity range gave mainly positive MN responses. Therefore, using RICC and RPD to select concentrations for MN analysis reduces the number of misleading positive results, regardless of cell type, by selecting lower concen- trations for analysis than when RCC or RI are used. The use of RICC and RPD does not result in a lower sensitivity of the assay. Others have shown that when RPD and RICC were used as measures of cytotoxicity of 14 known genotoxic agents, all selected concentra- tions for MN analysis gave rise to expected positive responses in a range of commonly used cell types (Kirkland, 2010). Another con- sideration is the effect of apoptosis inducing chemicals, such as curcumin and ethyl acrylate, which could contribute to the positive MN responses. In p53-competent TK6 cells, concentrations se- lected for MN analysis using the RCC and RI toxicity measures also caused increased levels of caspase activity, suggesting apoptosis had occurred. By contrast, when RPD and RICC were used to select concentrations, caspase levels were not significantly elevated, thus apoptosis was avoided and MN frequencies were close to back- ground levels (seeFowler et al., 2012bfor more detail on all of the results discussed above). It was concluded that a combination of careful selection of the cell type and toxicity measurement can significantly increase the Fig. 1.A comparison of the formation of MN in different cell types. Each cell type was tested with solvent (h) and low ( ), medium ( ) and high (j) concentrations of resorcinol for 3 h in the presence of S9. Resorcinol caused statistically significant increases (denoted as ‘‘ ’’) in MN formation in V79, CHL and CHO cells but not in human lymphocytes (HuLy), TK6 cells or HepG2 cells. (Reproduced fromFowler et al., 2012a). S. Pfuhler et al. / Toxicology in Vitro 28 (2014) 18–2319 specificity of clastogenicity assays (see also conclusions from an IWGT workshop reviewed byPfuhler et al., 2011). The now com- pleted ‘‘False Positives’’ project has undoubtedly lead to an improvement in the predictive capacity of thein vitroMNT. The higher predictive power of thein vitroassays will mean that fewer promising chemicals will be dropped from development without compromising safety of these ingredients. 3. The ‘‘3D skin model’’ project The first site of exposure to many cosmetic ingredients is the skin and, as a result, these chemicals may not enter the systemic circula- tion due to the barrier properties of the stratum corneum, or may be metabolized in the skin prior to entering the systemic circulation. Therefore, the most relevant model in which to test these chemicals is human skin. However, the availability of freshex vivohuman skin is sporadic which, together with the donor variability and differ- ences in tissue quality, makes the use of this model impractical for routine testing. Alternatives to native human skin are RS models which are prepared from primary human keratinocytes. These have a structure (Ponec et al., 2001) and metabolic capacity (Luu-The et al., 2009; Hu et al., 2010; Götz et al., 2012a,b; van Eijl et al., 2012) similar to that of native skin, making RS models a relevant and predictive model with which to test the genotoxic potential of dermally exposed chemicals. Indeed, the use of RS models in geno- toxicity risk assessment has been described recently (Pfuhler et al., 2010). Thein vitrohuman reconstructed skin micronucleus (RSMN) and Comet assays were developed for evaluating the genotoxicity of dermally applied chemicals. The project has been run according to a modular approach: Phase 1: Optimization and transferability of method across dif- ferent laboratories. Phase 2: Inter- and intra-laboratory reproducibility. Phase 3: Increasing the domain of chemicals tested for predic- tive capacity and further evaluation of reproducibility. The two endpoints have so far produced very promising results, as described below and summarized inTable 1. 4. The reconstructed skin micronucleus (RSMN) assay Thein vitroMN assay was adopted for use with RS models to as- sess the genotoxic potential of a number of chemicals selected by an independent Chemical Selection Expert team, including positive genotoxins with different mechanisms of action, true negatives and misleading positives (Kirkland et al., 2008). The initial proto- cols were defined (Curren et al., 2006) and then tested in three US laboratories (Mun et al., 2009; Hu et al., 2009). In Phase 1, the RSMN assay method was transferred to Henkel and L’Oréal, both European-based (Germany and France, respectively), which was of significance because the RS models were supplied from a US pro- vider (MatTek, MA). As part of this process, two training workshops were held to standardize the protocol and harmonize scoring of micronuclei, both of which were subsequently described and pub- lished byDahl et al. (2011). The workshops outlined three key is- sues impacting on the assay, namely the shipping (which should be overnight and under cooled conditions); the solvents used (avoid those that interfere with the air–liquid interface of the Epi- Derm™ model); and the subjective nature of the scorer (solved by creating a scoring atlas described byDahl et al., 2011). In Phase 2, three coded compounds (N-ethyl-N-nitrosourea (ENU), MMC (both genotoxic carcinogens) and cyclohexanone (non-carcino- gen and non-genotoxic)) were tested by three laboratories (Aardema et al., 2010). In addition to a good reproducibility between experi- ments within each laboratory, there was also a good inter-laboratory reproducibility for all three chemicals tested. Moreover, the genotoxic activity of each chemical was correctly identified in each laboratory. In Phase 3, the number of coded chemicals was increased to 29 as part of the validation process. All results were sent to ECVAM for decoding and evaluation according to specific pre-determined criteria. Results demonstrated an excellent specificity such that Fig. 2.Concentrations of misleading positive compounds resulting in 50–60% toxicity in TK6 cells according to different measurements. RPD and RICC tended to result in lower concentrations causing 50–60% toxicity and subsequently, fewer misleading positive responses. Bars with ‘‘+ve MN’’ indicate that the concentration indicated caused a statistically significant increase in MN formation (Reproduced fromFowler et al., 2012b). Table 1 Summary of the validation of the RS Comet and RSMN assays. RSMN assay RS Comet assay Phase 1Completed: Optimization of incubation conditions. Correct prediction of 5 dermal non- carcinogens and 7 model genotoxins including genotoxic dermal carcinogensCompleted: Assay readily adapted and transferred to different laboratories. Good intra- and inter-laboratory reproducibility with 2 model genotoxins Phase 2Completed: Transfer to different laboratories. Good intra- and inter-laboratory reproducibility of MN responses to 3 different coded chemicals. Correct identification of positive and negative genotoxicantsCompleted: Good intra- and inter-laboratory reproducibility of responses to 5 different coded chemicals. Correct identification of positive and negative genotoxicants. Optimization of RS model transport Phase 3On-going: Number of chemicals tested increased to 29. Initial results show improved specificity of the RSMN assay. Requires testing of additional positive chemicals to confirm sensitivity of the assayNot started. Project is merging with a project sponsored by the ‘‘Bundesministerium für Bildung und Forschung’’ (BMBF) to more efficiently test the validation chemicals 20S. Pfuhler et al. / Toxicology in Vitro 28 (2014) 18–23 approximately 90% of the experiments predictedin vivonon-geno- toxic non-carcinogens correctly (Fautz et al., 2012). Of the 29 chemicals tested, only 8 chemicals were classified as carcinogens with a suggested genotoxic mode of action and 21 were non-car- cinogens. Therefore, the current dataset is biased towards non-car- cinogens with the total number of carcinogens in the dataset considered too low to draw a final conclusion about the sensitivity of the RSMN assay. More coded compounds will be tested in the next project phase with a focus on carcinogens. During the testing, it was noted that a couple of compounds precipitated, which was not considered in the initial criteria, but should be included in future evaluations. A misleading negative may arise if the intended concentration is not reached or false pos- itives can be caused by precipitation due to practical issues such as difficulties in scoring (especially with compounds that fluoresce at the same wavelengths as acridine orange) or disruption of the air- liquid interface which can cause MN formation (Dahl et al., 2011). In order to speed up the scoring of the MN, efforts towards auto- mation are on the way. This should enable analysis of a greater number of cells, resulting in a higher statistical power of the assay and incorporate an automatic cytotoxicity measurement as part of the analysis. Initial results show good comparisons between man- ual scoring and flow cytometric methods (unpublished data). Some genotoxins require metabolic activation; therefore, we have investigated a number of chemicals that fall into this cate- gory, namely 4-nitroquinoline-n-oxide (4NQO), cyclophospha- mide, dimethylbenzanthracene (DMBA), dimethylnitrosamine (DMN), dibenzanthracene (DBA) and benzo[a]pyrene (BaP). Since the skin has been shown to have a very low phase 1 (normally bioactivating) capacity, it was considered that these chemicals may require a longer incubation duration in order to generate sufficient levels of the ultimate genotoxin. However, extending the dosing regimen from 48 h and two doses to 72 h and three doses did not always change the outcome of the assay, such that cyclophosphamide and DMBA were positive and DBA and DMN were negative using both dosing regimen. The outcome of the as- say was only changed for 4NQO, which was negative in the stan- dard 48 h dosing regimen, but positive with the 72 h treatment (Aardema et al., 2012). BaP gave mixed results, possibly due to this chemical precipitating at high concentrations and to altera- tions in metabolic enzyme levels caused by BaP (Götz et al., 2012c). The results for DBA and DMN may be expected since, for DBA, the level of CYP1A1/2 (which is needed for bioactivation of this molecule) is very low in both native skin and RS models (including Episkin and Epiderm models (van Eijl et al., 2012; Hu et al., 2010) such that the initial activation step is not possible or efficient enough in this organ. The low CYP1A1/2 expression is common to a number of RS skin models and to native human skin (van Eijl et al., 2012; Götz et al., 2012a) and is not a shortcoming of the EpiDerm model or 3D cultures in general. The lack of a genotoxic effect by DMN is reflected in the cancer bioassay, in which this compound did not cause tumours in the skin after application to the skin of rats (although it did result in tumours in the liver, lung and kidney) (Benemanskiı˘ and Levina, 1985). This suggests that DMN is not bioactivated in the skin or, at least to a sufficient extent to cause MN. Based on the result observed for 4NQO, it is recommended that a 48 h treatment is used for general testing and a longer treatment period is used when the outcome of the standard 48 h treatment is negative or question- able. This practice would not cost extra time, since a negative re- sult in the first 48 h experiment would be repeated using the 72 h dosing regimen instead of 48 h. These data support the conclusion that the RSMN assay in Epi- Derm™ is a valuablein vitromethod for genotoxicity assessment of dermally applied chemicals. The above mentioned global valida- tion project sponsored by Cosmetics Europe and ECVAM is contin- uing to collect data with a goal of demonstrating the strengths and limitations of this method. 5. The reconstructed skin Comet assay As with the RSMN assay, the RS Comet project was based on EpiDerm™ tissues. In contrast to the RSMN assay, in which the compounds are applied for at least 48 h, this assay involves the topical exposure to the chemicals for at least 3 h, followed by cell isolation and assessment of DNA damage using the Comet assay. Phase 1 studies have been completed. The method using 3D Epi- Derm™ tissues was readily adapted and transferred to different laboratories, showing good intra- and inter-laboratory reproduc- ibility with two model genotoxins, methyl methane sulfonate (MMS) and 4-NQO, in accordance within vivodata. In Phase 2, all laboratories tested five coded chemicals, and the reproducibility with these chemicals was generally good (Reuss et al., 2013). Phase 3 of this project has not yet started as we are still explor- ing opportunities to optimize the assay. Considerable intra- and in- ter-experimental variability was observed in some experiments, in which the solvent control values were greater than 30% (measured as % tail DNA). This variability was not attributable to a single fac- tor but was thought to be a stress-induced negative impact of transport on the quality of the tissues. The most promising way forward along with the below described use of full thickness mod- els is the use of ‘‘underdeveloped’’ (EPI-201) skin tissues rather than the standard MatTek epidermal (EPI-200) tissues. The under- developed EPI-201 tissues consist of the same normal, human-de- rived epidermal keratinocytes and are cultured in the same manner as standard EPI-200 tissues, but the underdeveloped tis- sues are shipped 4 days earlier than the normal tissues (on the week preceding the Comet assay). The EPI-201 tissues are then cul- tured further for 4 days in the receiving laboratory to produce tis- sues which are equivalent to the normal EPI-200 tissues. The additional time in culture after shipping may allow the tissues to recover from the transport-induced stress effects. The use of underdeveloped tissues changed the rate of invalid experiments due to a high background (>30% tail DNA in the untreated or vehi- cle treated group) in one laboratory from 4 in 8 experiments to 1 in 9 experiments. Despite the use of these tissues lowering the back- ground, the response of the tissues to the positive control, MMS, was unaltered, suggesting that this modification has not decreased the sensitivity of the assay (Reuss et al., 2013). The Cosmetics Europe RS Comet project is merging with a project sponsored by the German Federal Ministry of Education and Re- search (‘‘Bundesministerium für Bildung und Forschung’’, BMBF) in order to combine efforts and more efficiently test the validation chemicals in Phase 3. In the combined project we will be investigat- ing the usefulness of full-thickness skin models, aside from EPI 200/ 201, since they have been shown to have a higher metabolic capacity than epidermal models (Jäckh et al., 2011) and may better suited to address the genotoxic properties of pro-mutagens. 6. The ‘‘metabolism’’ project This project investigated xenobiotic metabolising enzymes (XMEs) in native human skin, RS and monolayer cultures of skin cells using both a proteomic approach and measurement of substrate metabolism. The proteomic methods included immunoblotting and a technique involving LC-MS/MS analysis of peptides and subse- quent software analysis (van Eijl et al., 2012). The latter method has the advantage of a much higher sensitivity than traditional immuno- blotting techniques and it allows for a comprehensive analysis of over 2000 XMEs. Phase 1 and 2 XME activities were measured using enzyme-selective substrates for cytochrome P450s (CYPs), sul- S. Pfuhler et al. / Toxicology in Vitro 28 (2014) 18–23 21 fotransferases (SULTs) and UDPGA-glucuronosyltransferases (UGTs) and glutathione S-transferases (GSTs) (Götz et al., 2012a, b,c). CYP 1–3 family proteins were not detected in native whole hu- man skin or any of thein vitromodels tested, which reflects the low mRNA expression of these CYPs reported by others (Luu-The et al., 2009) and the low or lacking metabolism of CYP-selective substrates in our studies (Götz et al., 2012a). The abundance of CYP1–3 proteins in human skin was estimated to be at levels at least 300-fold lower than that of liver. However, there were multiple other phase 1 XME proteins that were present in significant levels, such as alcohol dehydrogenases, aldehyde dehydrogenases, amine oxidases and epoxide hydrolases. GST proteins were the most abundant of the phase 2 enzymes investigated, and were present in both native hu- man skin and EpiDerm™ models. GST Pi was also identified as the most abundant isoform (van Eijl et al., 2012), which correlates with the high mRNA expression of this enzyme (Luu-The et al., 2009; Hu et al., 2010). The GST substrate, CDNB (Sherratt and Hayes, 2002), was metabolised at appreciable rates in whole skin ( 20 nmol/ min/mg), although this is still lower than that reported to be present in human liver (van Eijl et al., 2012; Götz et al., 2012b). The potential routes of metabolism in human skin and liver, based on their proteo- mic XME profiles, are depicted inFig. 3. The overall results from this project supported the view that skin tended to be more of a detoxi- fication than a bioactivation organ (in contrast to the liver) and that the levels of XMEs were all generally much lower than the liver. The XME profiles, using Affymetrix gene analysis, of different donors of EpiDerm™ models has been reported to be very similar (Hu et al., 2010) but there are no reports on how and if XMEs change during the course of an assay. Therefore, measurement of XMEs was adapted to determine how genotoxic compounds affect XME activities in EpiDerm™ models used under the conditions of the RSMN assay (Götz et al., 2012c). The genotoxins used were BaP and cyclophosphamide, both of which require bioactivation to their ultimate genotoxic metabolites. BaP caused a marked in- crease in ethoxyresorufinO-deethylase (reflecting CYP1A and CYP1B activity, also responsible for the bioactivation of BaPShi- mada and Fujii-Kuriyama, 2004) activities even 24 h after the first dose; and these raised levels continued until the final dose was applied (three doses in total and 72 h of exposure). Since the CYP1A/1B pathway was significantly increased, this would suggest that BaP may well cause an increase in MN formation in the Epi- Derm™ models in the RSMN assay. This was indeed the case for two experiments but not a third (Aardema et al., 2012), suggesting that there is an even balance of bioactivation via CYPs and detox- ification by GSTs (which were not induced by BaP) which can be tipped either way in different experiments. Unlike BaP, cyclophos- phamide did not alter any of the enzymes measured (Götz et al., 2012c), despite their involvement in its metabolism. Unlike BaP, cyclophosphamide was consistently positive in all experiments in both laboratories that tested this compound (Aardema et al., 2012). This suggests that the balance of metabolism of cyclophos- phamide in EpiDerm™ models is towards bioactivation. Overall, the measurement of certain XME activities in EpiDerm™ models treated under the conditions of an endpoint assay helped in the interpretation of the outcome. 7. Summary The Cosmetics Europe Genotoxicity Task Force projects, which have been running over the course of the last five years, have helped improve predictive capacity ofin vitroclastogenicity assays, and re- sulted in an increased understanding of the methods used to predict thein vivogenotoxic potential of dermally applied chemicals. Important protocol modifications, namely choice of the cell type and cytotoxicity measurement, have resulted in improved specific- ity of the MN test such that over 60% irrelevant positive findings could be prevented by using the optimized methods. This in itself will lead to an increase of the predictive capacity of the initial test battery ofin vitrotests and therefore reduce the number of chemi- cals de-selected due to misleading positives. The ‘‘3D skin model’’ project has shown that genotoxic endpoints, such as the MN and Co- met assays, can be adapted to RS models and that the protocols developed are robust, as demonstrated by the high degree of repro- ducibility between and within laboratories when coded compounds were tested. By establishing good predictivity of the RS MN and Co- met models, together with the confirmation that the RS models mi- mic native human skin in terms of their metabolic capacity (demonstrated in the ‘‘Skin Metabolism project’’), our results will support their use in follow-up tests in the assessment of the geno- toxic hazard of cosmetic ingredients in the absence ofin vivodata. Fig. 3.Potential routes of xenobiotic metabolism in skin and liver. The size of each arrow is proportional to the number of XMEs detected that may catalyze each bioconversion indicated. (Taken fromvan Eijl et al., 2012with kind permission from PLoS ONE,http://dx.doi.org/10.1371/journal.pone.0041721.g004.) 22S. Pfuhler et al. / Toxicology in Vitro 28 (2014) 18–23 Conflict of interest There are no conflicts of interests for any of the authors. Acknowledgements This work was funded by the Cosmetics Europe. The ‘‘3D skin model’’ project received input from ECVAM. Work by Covance and TNO was co-funded by ECVAM and UK NC3Rs. We would like to thank David Kirkland and Paul Fowler for their dedicated contri- butions to the ‘‘False Positives’’ project. We would also like to thank Marilyn Aardema for her role in the RSMN project and Raffa- ella Corvi for her help with steering the program. References Aardema, M.J., Barnett, B.C., Khambatta, Z., Reisinger, K., Ouedraogo-Arras, G., Faquet, B., Ginestet, A.C., Mun, G.C., Dahl, E.L., Hewitt, N.J., Corvi, R., Curren, R.D., 2010. International prevalidation studies of the EpiDerm 3D human reconstructed skin micronucleus (RSMN) assay: transferability and reproducibility. Mutat. Res. 701 (2), 123–131 . Aardema, M.J., Barnett, B., Mun, G., Dahl, E., Curren, R., Hewitt, N.J., Pfuhler, S., 2012. Evaluation of chemicals requiring metabolic activation in the EpiDermTM 3D human reconstructed skin micronucleus (RSMN) assay. Rev. Version Mut. Res . Benemanskiı˘ , V.V., Levina, VIa., 1985. Carcinogenic effect of N- nitrosodimethylamine after application to rat skin. Eksp. Onkol. 7 (2), 20–21 . Curren, R.D., Mun, G.C., Gibson, D.P., Aardema, M.J., 2006. Development of a method for assessing micronucleus induction in a 3D human skin model (EpiDerm). Mutat. Res. 607 (2), 192–204 . Dahl, E.L., Curren, R., Barnett, B.C., Khambatta, Z., Reisinger, K., Ouedraogo, G., Faquet, B., Ginestet, A.C., Mun, G., Hewitt, N.J., Carr, G., Pfuhler, S., Aardema, M.J., 2011. The reconstructed skin micronucleus assay (RSMN) in EpiDerm™: detailed protocol and harmonized scoring atlas. Mutat. Res. 720 (1–2), 42–52 . Erexson, G.L., Periago, M.V., Spicer, C.S., 2001. Differential sensitivity of Chinese hamster V79 and Chinese hamster ovary (CHO) cells in the in vitro micronucleus screening assay. Mutat. Res. 495 (1–2), 75–80 . EU, 2003, EC – Directive 2003/15/EC of the European parliament and of the council of 27 February 2003 amending council directive 76/768/EEC on the approximation of the laws of the member states relating to cosmetic products. Official Journal L66, 11/03/2003, p. 26. European Commission, 2006. Regulation (EC) No. 1907/2006 of the European parliament and of the council of 18 December 2006 concerning the REGISTRATION, EValuation, authorisation and restriction of chemicals (REACh), establishing a European chemicals agency, amending directive 1999/ 45/EC and repealing council regulation (EEC) No. 793/93 and commission regulation (EC) No. 1488/94 as well as council directive 76/769/EEC and commission directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC. http://eur-lex.europa.eu/LexUriServ/ LexUriServ.do?uri=CELEX:32006R1907:EN:NOT. Fautz, R., Curren, R., Krul, C., Reisinger, K., Ouedraogo, G., Corvi, R., Aardema, M., Reus, A., Barnett, B., Downs, T., Faquet, B., Hoffmann, S., Hewitt, N., Barroso, J., Pfuhler, S. in: Poster presentation at the EPAA Annual Conference 2012 – Global Cooperation on alternatives (3Rs) to animal testing, Brussels, November, 2012: Pre-validation of the Reconstructed 3D Human Skin Micronucleus and Comet Assay. http://ec.europa.eu/enterprise/epaa/3_events/ann-conf-2012/poster- book.pdf. Fellows, M.D., O’Donovan, M.R., Lorge, E., Kirkland, D., 2008. Comparison of different methods for an accurate assessment of cytotoxicity in the in vitro micronucleus test. II: Practical aspects with toxic agents. Mutat. Res. 655 (1–2), 4–21 . Fowler, P., Smith, K., Young, J., Jeffrey, L., Kirkland, D., Pfuhler, S., Carmichael, P., 2012a. Reduction of misleading (‘‘false’’) positive results in mammalian cell genotoxicity assays. I. Choice of cell type. Mutat. Res. 742 (1–2), 11–25 . Fowler, P., Smith, R., Smith, K., Young, J., Jeffrey, L., Kirkland, D., Pfuhler, S., Carmichael, P., 2012b. Reduction of misleading (‘‘false’’) positive results in mammalian cell genotoxicity assays. II. Importance of accurate toxicity measurement. Mutat. Res. 747 (1), 104–117 . Fowler, P., Smith, R., Smith K., Young, J., Jeffrey, L., Kirkland, D., Pfuhler, S., Carmichael, P., 2013. Reduction of misleading (‘‘false’’) positive results in mammalian cell genotoxicity assays. III. Sensitivity of human cell types, Mutat Res., accepted for publication. Fowler, P., Smith, R., Smith, K., Young, J., Jeffrey, L., Kirkland, D., Pfuhler, S., Carmichael, P., 2013. Reduction of misleading (‘‘false’’) positive results in mammalian cell genotoxicity assays. Accepted for publication in Mutation Research, III. Sensitivity of human cell types . Götz, C., Pfeiffer, R., Tigges, J., Hübenthal, U., Ruwiedel, K., Freytag, E.-M., Merk, H.F., Krutmann, J., Edwards, R.J., Abel, J., Pease, C., Goebel, C., Hewitt, N.J., Fritsche, E., 2012a. Xenobiotic metabolism capacities of human skin in comparison to 3D- epidermis models and keratinocyte-based cell culture as in vitro alternatives for chemical testing: phase I. Exp. Dermatol. 21 (5), 358–363 . Götz, C., Pfeiffer, R., Tigges, J., Hübenthal, U., Ruwiedel, K., Freytag, E.-M., Merk, H.F., Krutmann, J., Edwards, R.J., Abel, J., Pease, C., Goebel, C., Hewitt, N.J., Fritsche, E., 2012b. Xenobiotic metabolism capacities of human skin in comparison to 3D- epidermis models and keratinocyte-based cell culture as in vitro alternatives for chemical testing: phase 2. Exp. Dermatol. 21 (5), 364–369 . Götz, C., Hewitt, N.J., Jermann, E., Tigges, J., Kohne, Z., Hübenthal, U., Krutmann, J., Merk, H.F., Fritsche, E., 2012c. Effects of the genotoxic compounds, benzo[a]pyrene and cyclophosphamide on phase 1 and 2 activities in EpiDerm™ models. Xenobiotica 42 (6), 526–537 . Hilliard, C., Hill, R., Armstrong, M., Fleckenstein, C., Crowley, J., Freeland, E., Duffy, D., Galloway, S.M., 2007. Chromosome aberrations in Chinese hamster and human cells: a comparison using compounds with various genotoxicity profiles. Mutat. Res. 616 (1–2), 103–118 . Hu, T., Kaluzhny, Y., Mun, G.C., Barnett, B., Karetsky, V., Wilt, N., Klausner, M., Curren, R.D., Aardema, M.J., 2009. Intralaboratory and interlaboratory evaluation of the EpiDerm 3D human reconstructed skin micronucleus (RSMN) assay. Mutat. Res. 673 (2), 100–108 . Hu, T., Khambatta, Z.S., Hayden, P.J., Bolmarcich, J., Binder, R.L., Robinson, M.K., Carr, G.J., Tiesman, J.P., Jarrold, B.B., Osborne, R., Reichling, T.D., Nemeth, S.T., Aardema, M.J., 2010. Xenobiotic metabolism gene expression in the EpiDermin vitro 3D human epidermis model compared to human skin. Toxicol. In Vitro 24 (5), 1450–1463 . Jäckh, C., Blatz, V., Fabian, E., Guth, K., van Ravenzwaay, B., Reisinger, K., Landsiedel, R., 2011. Characterization of enzyme activities of Cytochrome P450 enzymes, Flavin-dependent monooxygenases, N-acetyltransferases and UDP- glucuronyltransferases in human reconstructed epidermis and full-thickness skin models. Toxicol. In Vitro 25 (6), 1209–1214 . Kirkland, D., Aardema, M., Henderson, L., Müller, L., 2005. Evaluation of the ability of a battery of three in vitro genotoxicity tests to discriminate rodent carcinogens and non-carcinogens I. Sensitivity, specificity and relative predictivity. Mutat. Res. 584 (1–2), 1–256 . Kirkland, D., Pfuhler, S., Tweats, D., Aardema, M., Corvi, R., Darroudi, F., Elhajouji, A., Glatt, H., Hastwell, P., Hayashi, M., Kasper, P., Kirchner, S., Lynch, A., Marzin, D., Maurici, D., Meunier, J.R., Müller, L., Nohynek, G., Parry, J., Parry, E., Thybaud, V., Tice, R., van Benthem, J., Vanparys, P., White, P., 2007. How to reduce false positive results when undertaking in vitro genotoxicity testing and thus avoid unnecessary follow-up animal tests: report of an ECVAM Workshop. Mutat. Res. 628 (1), 31–55 . Kirkland, D., Kasper, P., Müller, L., Corvi, R., Speit, G., 2008. Recommended lists of genotoxic and non-genotoxic chemicals for assessment of the performance of new or improved genotoxicity tests: a follow-up to an ECVAM workshop. Mutat. Res. 653 (1–2), 99–108 . Kirkland, D., 2010. Evaluation of different cytotoxic and cytostatic measures for the in vitro micronucleus test (MNVit): summary of results in the collaborative trial. Mutat. Res. 702 (2), 139–147 . Luu-The, V., Duche, D., Ferraris, C., Meunier, J.R., Leclaire, J., Labrie, F., 2009. Expression profiles of phases 1 and 2 metabolizing enzymes in human skin and the reconstructed skin models Episkin and full thickness model from Episkin. J. Steroid Biochem. Mol. Biol. 116 (3–5), 178–186 . Mun, G.C., Aardema, M.J., Hu, T., Barnett, B., Kaluzhny, Y., Klausner, M., Karetsky, V., Dahl, E.L., Curren, R.D., 2009. Further development of the EpiDerm 3D reconstructed human skin micronucleus (RSMN) assay. Mutat. Res. 673 (2), 92–99 . OECD, 2010. OECD guideline for the testing of chemicals draft proposal for a new guideline 487: in vitro micronucleus test. (adopted 22.07.10).http:// iccvam.niehs.nih.gov/SuppDocs/FedDocs/OECD/OECD-TG487.pdf. Pfuhler, S., Kirst, A., Aardema, M., Banduhn, N., Goebel, C., Araki, D., Costabel-Farkas, M., Dufour, E., Fautz, R., Harvey, J., Hewitt, N.J., Hibatallah, J., Carmichael, P., Macfarlane, M., Reisinger, K., Rowland, J., Schellauf, F., Schepky, A., Scheel, J., 2010. A tiered approach to the use of alternatives to animal testing for the safety assessment of cosmetics: genotoxicity. A COLIPA analysis. Regul. Toxicol. Pharmacol. 57 (2–3), 315–324, Erratum in: Regul. Toxicol. Pharmacol. 58(3), 544 . Pfuhler, S., Fellows, M., van Benthem, J., Corvi, R., Curren, R., Dearfield, K., Fowler, P., Frötschl, R., Elhajouji, A., Le Hégarat, L., Kasamatsu, T., Kojima, H., Ouédraogo, G., Scott, A., Speit, G., 2011. In vitro genotoxicity test approaches with better predictivity: summary of an IWGT workshop. Mutat. Res. 723 (2), 101–107 . Ponec, M., Gibbs, S., Pilgram, G., Boelsma, E., Koerten, H., Bouwstra, J., Mommaas, M., 2001. Barrier function in reconstructed epidermis and its resemblance to native human skin. Skin Pharmacol. Appl. Skin Physiol. 14 (Suppl. 1), 63–71 . Reuss, A., Reisinger, R., Downs, T.R., Carr, G., Zeller, A., Corvi, R., Krul C.A.M., Pfuhler, S., 2013. Comet assay in reconstructed 3D human epidermal skin models – investigation of intra- and inter-laboratory reproducibility with coded chemicals, Mutagenesis, accepted for publication. SCCP, 2009. SCCP/1212/09. Position statement on genotoxicity/mutagenicity testing of cosmetic ingredients without animal experiments.http://ec.europa.eu/ health/ph_risk/committees/04_sccp/docs/sccp_s_08.pdf. Sherratt, P.J., Hayes, J.D., 2002. Glutathione S-transferases. In: Ioannides, C. (Ed.), Enzyme Systems that Metabolise Drugs and Other Xenobiotics. Wiley & Sons, Chichester, UK, pp. 319–352 . Shimada, T., Fujii-Kuriyama, Y., 2004. Metabolic activation of polycyclic aromatic hydrocarbons to carcinogens by cytochromes P450 1A1 and 1B1. Cancer Sci. 95 (1), 1–6 . Sofuni, T., Matsuoka, A., Sawada, M., Ishidate Jr., M., Zeiger, E., Shelby, M.D., 1990. A comparison of chromosome aberration induction by 25 compounds tested by two Chinese hamster cell (CHL and CHO) systems in culture. Mutat. Res. 241 (2), 175–213 . van Eijl, S., Zhu, Z., Cupitt, J., Gierula, M., Götz, C., Fritsche, E., Edwards, R.J., 2012. Elucidation of xenobiotic metabolism pathways in human skin and human skin models by proteomic profiling. PLoS One 7 (7), e41721 . S. Pfuhler et al. / Toxicology in Vitro 28 (2014) 18–2323
I need to write 5 bibliographies for an essay I’m doing discussing how we shouldn’t use animals for cosmetic testing. I included the 5 sources and 2 files that explain how it should be done and an exa
Do Animals Feel Pain? Author(s): Peter Harrison Source: Philosophy , Jan., 1991 , Vol. 66, No. 255 (Jan., 1991), pp. 25-40 Published by: Cambridge University Press on behalf of Royal Institute of Philosophy Stable URL: https://www.jstor.org/stable/3751139 JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms and Cambridge University Press are collaborating with JSTOR to digitize, preserve and extend access to Philosophy This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? PETER HARRISON In an oft-quoted passage from The Principles of Morals and Legislation (1789yf – H U H P % H Q W K D P D G G U H V V H V W K H L V V X H R I R X U W U H D W P H Q W R I D Q L – mals with the following words: ‘the question is not, Can they reason? nor, can they talk? but, Can they suffer?’1 The point is well taken, for surely if animals suffer, they are legitimate objects of our moral con- cern. It is curious therefore, given the current interest in the moral status of animals, that Bentham’s question has been assumed to be merely rhetorical. No-one has seriously examined the claim, central to arguments for animal liberation and animal rights, that animals actually feel pain. Peter Singer’s Animal Liberation is perhaps typical in this regard. His treatment of the issue covers a scant seven pages, after which he summarily announces that ‘there are no good reasons, scien- tific or philosophical, for denying that animals feel pain’.2 In this paper I shall suggest that the issue of animal pain is not so easily dispensed with, and that the evidence brought forward to demonstrate that ani- mals feel pain is far from conclusive. Three kinds of argument are commonly advanced to support the contention that animals feel pain. The first involves the claim that animal behaviours give us clues to alleged mental states, about what animals are feeling. Thus animals confronted with noxious stimuli which would cause human beings pain, react in similar ways. They attempt to avoid the stimulus, they show facial contortions, they may even cry out. From these ‘pain behaviours’ it is inferred that the animals must be experiencing pain. A second argument asserts that by virtue of a similarity in structure and function of nervous systems it is likely that human beings and animals closely related to the human species will experience the exter- nal environment in much the same way. It is assumed, for example, that primates have visual experiences similar to our own, feel hunger and thirst as we do, and so on. Presumably when they encounter noxious stimuli, they, like us, feel pain. A third line of argument is derived from evolutionary theory. Organic evolution implies that there is no radical discontinuity between human and other species. It is likely, on this view, that human minds 1 (Oxford: Clarendon Press, 1907yf I Q ; 9 , , L Y f. 2 Peter Singer, Animal Liberation (London: Cape, 1976yf . Philosobhv 66 1991 25This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison evolved from animal minds, and that closely related species would experience similar mental events. The evolutionary model would also suggest that pain is an essential adaptation for organisms in that it helps them avoid those things which would reduce their chances of survival and reproduction. Let us consider these arguments in turn. I The argument based on ‘pain behaviours’ is the most intuitive. Con- sidered in isolation, however, it is the least compelling. Even the simplest representatives of the animal kingdom exhibit rudimentary ‘pain behaviours’. Single-celled organisms, for example, will withdraw from harmful stimuli. Insects struggle feebly after they have been inadvertently crushed underfoot. Yet few would want to argue that these behaviours resulted from the experience of pain. Certainly we show little sympathy for those unfortunate ants which are innocent casualties of an afternoon stroll, or the countless billions of micro- organisms destroyed by the chlorination of our water supplies. For all practical purposes we discount the possibility that such simple forms of life feel pain, despite their behaviours. In more elevated levels of the animal kingdom there are also instances of ‘pain behaviours’ which undoubtedly occur in the absence of pain. Some parent birds, for instance, will feign injury to lure predators away from their young. The converse is also true. Animals might have sustained considerable tissue damage, but display none of the signs which we imagine would usually attend such trauma. This is because immobility is the best response to certain kinds of injury.3 Pain behaviours, in any case, can be ably performed by non-living entities. If we were to construct a robot which was devoid of speech, yet was to have an active and independent existence, it would be necessary to programme it with mechanisms of self-preservation. Of the many objects it might encounter, it would need to be able to detect and 3 Thus Dennis and Melzack: ‘The appropriate behavioural response to overt damage may be inactivity; pain arising from trauma should presumably promote such behaviour. However, the appropriate behavioural response to threat may be vigorous activity; pain arising from threat should therefore promote this sort of activity. Thus the overt expression of pain sensation may actually be a combination of inherently contradictory processes and behavioural tendencies.’ S. Dennis and R. Melzack, ‘Perspectives on Phy- logenetic Evolution of Pain Expression’, Animal Pain: Perception and Allevi- ation, R. L. Kitchell and H. H. Erickson (edsyf % H W K H V G D $ P H U L F D n Physiological Society, 1983yf . 26This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? respond to those likely to cause it most harm. Properly programmed, such a machine would manifest its own ‘pain behaviour’. If we lit a fire under it, it would struggle to escape. If it found itself in a dangerous situation from which it could not extricate itself (say it fell into an acid bathyf L W Z R X O G D W W H P S W W R V X P P R Q D L G Z L W K V K U L O O F U L H V , I L W Z H U e immobilized after a fall, it might, by facial contortions, indicate that it was damaged. But this ‘pain behaviour’ would convey nothing about what it was feeling, for robots, on most accounts, can feel nothing. All that could be learned from such behaviour was how well the robot had been programmed for self-preservation. Mutatis mutandis, the ‘pain behaviours’ of animals demonstrate, in the first instance, how well natural selection has fitted them for encounters with unfriendly aspects of their environment. For neither animals, nor our imaginary robot, is ‘pain behaviour’ primarily an expression of some internal state. I think these examples are sufficient to show that the argument from behaviours alone is fairly weak. But the reason we are inclined to deny that simple animals and computers feel pain is that despite their compe- tent performance of ‘pain behaviours’, their internal structure is suffici- ently dissimilar to our own to warrant the conclusion that they do not have a mental life which is in any way comparable. Animals closely related to the human species, however, possess at least some of the neural hardware which in human beings is thought to be involved in the experience of pain. It might be that the behavioural argument is stronger when considered together with the second argument-that based on the affinity of nervous systems. II Pain is a mental state. It might be caused by, or correlated with, brain states. It might have behavioural or psychological indicators. Yet it remains intractably mental. Herein lies the stumbling block of the second argument, for the closest scrutiny of the nervous systems of human beings and animals has never progressed beyond, and arguably never will progress beyond, the description of brain states to arrive at mental states. Thus the introduction of the structure and function of nervous systems into this discussion brings with it that whole constella- tion of difficulties which revolve around the problem of psycho-physi- cal reductionism. Can mental states be reduced to physical states, and is it possible to project mental states from appropriate anatomical and physiological data? To be successful, the second argument for animal pain must answer both of these questions in the affirmative. Descartes, in his Meditations (1641yf T X L W H F R U U H F W O S R L Q W H G R X W W K D t there is no necessary logical relation between propositions about mental 27This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison states and propositions about physical states. We may doubt the exist- ence of our bodies, but not our minds. A disembodied mind is a logical possibility. Conversely, there is no logical impropriety in imagining bodies behaving in quite complex ways, without those behaviours being necessarily accompanied by relevant mental processes. Our robot, for example, would fit the bill, and indeed for Descartes, animals too were merely automatons, albeit organic ones. Of course from the fact that there is no logical connection between mental states and physical states it cannot be inferred that no contingent connection is possible. Descriptions of mental and physical states may be linked in a number of ways, and it is upon such linkages that the second argument for animal pain depends. The most compelling evi- dence of connection between the physical state of the brain and the mental life of the individual comes from instances of brain pathology or brain surgery. The fact that damage to the cerebral cortex can reduce individuals to a ‘mindless’ state would suggest that observable brain states cause mind states, or at the very least are a necessary condition of mind states. More specifically, neurologists have had some success in identifying those parts of the brain which seem to be responsible for particular conscious states. Our experience of pain, for example, seems to be mediated through a complicated physical network involving the neospinothalamic projec- tion system (sensory aspects of painyf U H W L F X O D U D Q G O L P E L F V W U X F W X U H s (motivational aspects of painyf D Q G W K H Q H R F R U W H [ R Y H U D O O F R Q W U R O R f sensory and motivational systemsyf , W P D E H V L J Q L I L F D Q W W K D W W K L s latter structure we share only with the primates. An argument could be made on this basis alone that the experience which we designate ‘pain’ is peculiar to us and a few primate species.yf % X W G H V S L W H V X F K Z H O O – established connections between observable brain structures and more elusive mental states, it would be rash to attempt to predict the mental states of individuals on the basis of the presence or absence of certain structures, or even on the basis of the physiological status of those structures.5 The well-known literature on the psychology of pain illus- trates that the same stimulus may prove intensely painful to one indi- vidual, and be of little concern to another. The use of placebos to 4 See, e.g., Ronald Melzack, The Puzzle of Pain (Ringwood: Penguin, 1973yf I . 5 Thus Theodore Barber reports of individuals chronically insensitive to pain that for most, if not all, ‘no distinct localized damage exists in the central nervous system’. ‘Toward a Theory of Pain’, Psychological Bulletin 56 (1959yf , 443. It is true that Barber cites no evidence from autopsies, and that more sophisticated scanning apparatus has been developed since this publication, but the fact that this insensitivity to pain can be reversed without surgical intervention would support Barber’s observation. 28This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? control pain, the influence of hypnosis or suggestion to influence pain perception, national differences in pain thresholds, all such aspects of the psychology of pain illustrate that the presence of certain brain structures and requisite sensory inputs are not sufficient conditions for the prediction of mental states. Not only does the psychology of pain afford instances in which the same neural hardware might give rise to a variety of different conscious states, but the human brain itself exhibits an amazing ability to generate certain mental states in the absence of the relevant physical structures. Phantom pain is perhaps the most obvious example. Amputees fre- quently report awareness of a limb which has been recently amputated. In a minority of cases a phantom limb may become an ongoing source of severe pain. Often the pain is located in a quite specific part of the missing appendage. An even more compelling illustration of the generation of certain mental states in the absence of appropriate structures comes from John Lorber’s engaging paper ‘Is Your Brain Really Necessary?’6 Paediatric neurologist Lorber reports on a number of individuals with hydro- cephalus-a condition which resulted in their having virtually no cere- bral cortex. The most intriguing case cited by Lorber is that of a mathematician with IQ of 126. A brain scan revealed that this young man had, in Lorber’s words, ‘virtually no brain’. The supratentorial part of the intracranial cavity contained only a thin layer of brain tissue, between one and two millimetres thick, attached to the skull wall. No ‘visual cortex’ was evident, yet the individual, who by all accounts should have been blind, had above average visual perception. It is likely that the functions which would normally have taken place in the missing cerebral cortex had been taken over by other structures. Cases such as this show that certain aspects of human consciousness have a tenacity which confounds our understanding of the link between brain structure and consciousness. Lorber’s discoveries are a striling example of the fact that an advanc- ing neuroscience, far from establishing concrete links between brain states and mental states, is actually deepening the mystery of how the brain is causally related to human consciousness. It need hardly be said that when we cross the species boundary and attempt to make projec- tions about animals’ putative mental lives based on the structures of their nervous systems we are in murky waters indeed. Two further examples illustrate this. 6 See David Paterson’s article of the same name in World Medicine 3 May 1980, 21-24. Also see Norton Nelkin, ‘Pains and Pain Sensations’, The Jour- nal of Philosophy 83 (1986yf . 29This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison The brains of birds, such as they are, do not contain a ‘visual cortex’. Thus if we are to argue that similar brain structures give rise to similar experiences, then it is unlikely that the visual experiences of birds will be qualitatively similar to our own. On the other hand, the behaviour of birds would seem to indicate that they can ‘see’. While we assume from the behaviour of birds that their visual experience of the world is much the same as ours, if we are committed to the view that like mental states are generated by like brain stuctures, we are bound to admit that this assumption is unfounded. We might of course be tempted to revert to the first argument-that behaviour, not structure, gives the correct cues to mental states. But this seems to commit us to the view that computers, flies, and amoebas have states of consciousness like our own. Another illustration which concerns visual experiences is the much- discussed phenomenon of ‘blind-sight’.7 As we have already men- tioned, the ‘visual’ or striate cortex is thought to be necessary for human vision. Individuals suffering from damage to the striate cortex may lose sight in part of their visual field. Larry Weisenkrantz and his colleagues have carried out a number of experiments on one such individual who claimed to be blind in his left field of view. Simple shapes were presented to this subject in his blind field of view. Though he denied being able to see anything, the subject could, with reasonable con- sistency, describe the shape of the object and point to it. In each instance he insisted that his correct response was merely a guess.8 Examples of blindsight indicate, amongst other things, that it is pos- sible to have visual experiences of which we are unaware. The blind- sight phenomenon thus opens up the possibility that there might be non-conscious experiences to which we can nonetheless respond with the appropriate behaviour.9 Blindsighted individuals can learn to respond as if they see, even though they have no conscious awareness of seeing anything. The significance of this for a discussion of animal behaviours is that animals might respond to stimuli as if they were conscious of them, while in fact they are not. Thus birds which lack the human apparatus of conscious vision (as do blindsighted subjectsyf might not simply have qualitatively different visual experiences as 7 On ‘blindsight’ see Larry Weisenkrantz, ‘Varieties of Residual Experi- ence’, Quarterly Journal of Experimental Psychology 32 (1980yf ; Thomas Natsoulas, ‘Conscious Perception and the Paradox of “Blindsight”‘, in Aspects of Consciousness, III, Geoffrey Underwood (ed.yf / R Q G R Q $ F D – demic Press, 1982yf . 8 Larry Weisenkrantz, ‘Trying to Bridge some Neurophysiological gaps between Monkey and Man’, British Journal of Psychology 68 (1977yf . 9 On the possibility of ‘non-conscious experience’, see Peter Carruthers, ‘Brute Experience’, The Journal of Philosophy 86 (1989yf . 30This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? suggested above, they might not have conscious visual experiences at all. It may be concluded that an animal’s experience of stimuli which we would find painful might be qualitatively different (that is, not painfulyf or may even be non-conscious. Animals might react to such stimuli by exhibiting ‘pain behaviour’ and yet not have that mental experience which we call ‘pain’, or perhaps not have any conscious experience at all.10 So far our discussion of neural circuitry and how it relates to putative mental states has focused upon the inability of contemporary neuro- science to bridge the gap between brain and mind. There are those, of course, who have asserted that it is impossible in principle to bridge that gap. It is significant that Thomas Nagel, one of the chief spokes- men for this group, has alluded to animal consciousness to make his point. In the seminal paper ‘What is it Like to be a Bat?’,1 Nagel leads us into the subjective world of the bat. These curious mammals, he reminds us, perceive the external world using a kind of sonar. By emitting high-pitched squeals and detecting the reflections, they are able to create an accurate enough image of their environment to enable them to ensnare small flying insects, while they themselves are air- borne. Nagel points out that we might observe and describe in detail the neurophysiology which makes all this possible, but that it is unlikely that any amount of such observation would ever give us an insight into the bat’s subjective experience of the world-into what it is like to be a bat. As Nagel himself puts it: For if the facts of experience-facts about what it is like for the experiencing organism-are available only from one point of view, then it is a mystery how the true character of experience could be revealed in the physical operation of that organism.’2 Nagel thus asserts that the construction of subjective experiences from the observation of brain states is in principle impossible.13 For our present purposes it is not necessary to enter into the argu- ment about whether mind states are reducible to brain states. Suffice it 10 This is also suggested by Carruthers, ibid., 266-269. 1 The Philosophical Review 83 (1974yf . 12 Ibid., 442. 13 Colin McGinn has made a similar point from a different perspective. He argues that the mystery of our mental life arises out of the fact that we simply do not possess the cognitive faculties necessary to solve the mind-body prob- lem. ‘Cognitive closure’ prevents our ever having access to that vital natural link which presumably exists between brain states and conscious states. See Colin McGinn, ‘Can We Solve the Mind-Body Problem?’, Mind 98 (1989yf , 349-366. 31This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison to say that there is sufficient confusion about how brain structure and function relate to mental states to rule out any simple assertion that animal nervous systems which resemble our own will give rise to mental states like ours. It seems then, that pain, a mental state, can be neither perceived nor inferred by directing the senses on to behaviours or on to the brain itself. But what of the third argument for animal pain-that based on evolutionary theory? III Evolutionary theory provides the most convincing case for animal pain. Because evolution stresses continuities in the biological sphere, it breaks down the distinction between human and animal. Thus any special claims made on behalf of the human race-that they alone experience pain, for example-require justification. Before examining how, in evolutionary terms, we might justify treating Homo sapiens as a unique case, we ought to consider first how animal pain might con- ceivably fit into the evolutionary scheme of things. Natural selection ‘designs’ animals to survive and reproduce. An important sort of adaptation for organisms to acquire would be the ability to avoid aspects of the environment which would reduce their chances of survival and reproduction. Pain, we might suppose, plays this adaptive role by compelling organisms to avoid situations in their world which might harm them. This view of the matter receives some measure of support from cases of individuals born with a congenital insensitivity to pain. Such unfortunate people frequently injure them- selves quite severely in their early childhood, and must be taught how to avoid inflicting damage upon themselves. That such a condition can lead eventually to permanent disability or death would suggest that pain has considerable adaptive value for human beings at least.14 Ani- mals which were similarly insensitive to damaging stimuli, we might reasonably infer, would have little chance of survival. Yet there are difficulties with this interpretation. Strictly, it is not pain (real or imputedyf Z K L F K L V W K H D G D S W D W L R Q E X t the behaviour which is elicited when the damaging stimulus is applied. Those who are insensitive to pain are not disadvantaged by the absence of unpleasant mental states, but by a lack of those behavioural responses which in others are prompted by pain. We tend to lose sight of the primacy of behaviour because we get caught up in the con- notations of ‘expression’. That is to say, we consider some animal 14 On congenital insensitivity to pain see Melzack, The Puzzle of Pain, 15f. 32This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? behaviours to be expressions of a particular mental state. Even Darwin, who should have known better, was guilty of this infelicity when he spoke of the ‘expression of the emotions in man and animals’. Such locutions are misleading because they suggest that certain aspects of animal behaviour are arbitrary outward signs which signify some con- scious state. But the simplest application of the theory of natural selection would only allow that such behaviours as violent struggling, grimacing and crying out, serve some more direct purpose in enhancing an animal’s chances of survival and reproduction. (Darwin admittedly stressed the communicative aspects of these signs.yf 7 R H [ S O R L W D Q R W K H r example which I have drawn upon in another context, a wildebeest which is being torn apart by dogs will die in silence, while a chimpanzee will screech out in response to some trivial hurt like a thorn puncturing its foot.15 It seems that the chimp gives expression to its pain, whereas the wildebeest does not. Yet neither expresses its pain. Rather, each behaves in a way likely to enhance the survival of the species. The chimpanzee communicates either to warn its conspecifics, or to sum- mon aid. The wildebeest remains silent so that others will not be lured to their deaths. It is the behaviour, rather than some hypothetical mental state, which adapts the organism. Another linguistic usage which holds us in thrall is the language of ‘detection’. We assume that ‘detection’ entails ‘conscious awareness of’. This leads us to believe that an animal cannot respond to a stimulus unless in some sense it consciously ‘knows’ what it has encountered. The reason such insectivorous plants as the venus fly trap capture our imagination is that they behave as if they are aware. How, we ponder, do they ‘know’ that the fly is there? Again we need to remind ourselves that the simplest of organisms are able to detect and respond to stimuli, yet we are not thereby committed to the view that they have knowledge or beliefs. The same is true of more neurologically complex organisms. There is an important truth in that litany of behaviourists: animals acquire behaviours, not beliefs. If it is granted that the behaviour rather than some postulated mental state is what adapts an organism, we are next led to inquire whether organisms might exhibit ‘pain behaviours’ without that attendant men- tal state which we call ‘pain’. As we noted at the outset, many invertebr- ates to which we do not generally attribute feelings of pain exhibit ‘pain behaviour’. In higher animals too, as we have already seen, it is possible that relevant behaviours might be performed in the absence of any conscious experience. But is it probable? Must pain be introduced to cause the behaviours, or might these be caused more directly by the 15 David McFarland, ‘Pain’, The Oxford Companion to Animal Behaviour, David McFarland (ed.yf 2 [ I R U G 8 Q L Y H U V L W 3 U H V V f, 439. 33This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison stimulus, or perhaps by indifferent conscious states? We might at this point simply opt for the most parsimonious explanation. This is in fact the upshot of Lloyd Morgan’s famous dictum: ‘In no case may we interpret an action as the outcome of the exercise of a higher psychical faculty, if it can be interpreted as the outcome of the exercise of one which stands lower in the psychological scale.’16 We must ask, in other words, if we can explain all animals’ reactions to noxious stimuli without recourse to particular mental states. Our blindsight examples show that it is possible for organisms to respond appropriately to stimuli in the complete absence of mental states. If the general case is true, then the same might be said for the specific performance of ‘pain behaviours’ in the absence of pain. The thrust of Morgan’s canon can be reinforced epistemologically with the arguments of Descartes. As we know, Descartes’ radical doubt led him to propose that all we can know for certain are the truths of logic and the existence of our own mental states.17 Fortunately one of the truths of logic was the existence of a God who could guarantee, to some extent, the veracity of perceptions of the world. Yet strict application of the criterion of doubt permits us to ascribe minds to other creatures only if they demonstrate (verbally, by signs, or by rational behaviouryf evidence of mental activity. From the lack of such indications from animals, Descartes concluded that we have no evidence which would enable us legitimately to infer that animals have minds.18 Not having minds, they cannot feel pain. Descartes thus provides epistemological grounds for denying that animals feel pain.19 If we adopt the conservative stance of Morgan or Descartes, then it seems that we have no grounds, scientific or philosophical, for asserting that animals feel pain. Yet this is a much weaker claim than the positive assertion that we have good reasons for believing that animals do not 16 Quoted in Robert Boakes, From Darwin to Behaviourism (Cambridge University Press, 1984yf 7 K L V G L F W X P L V D F W X D O O D Y H U V L R Q R I W K H $ U L V W R – telian principle, ‘Nature does nothing in vain’, couched in evolutionary terms. 17 Meditations II. 18 Descartes’ clearest explanation of the matter comes in a letter to the English Platonist, Henry More. See Descartes, Philosophical Letters, Anthony Kenny (ed.yf 2 [ I R U G & O D U H Q G R Q 3 U H V V f, 243-245. 19 It may seem that Morgan and Descartes are making the same point, but they are not. Morgan’s canon was virtually a biological application of the second law of thermodynamics, asserting that a complex biological system would not evolve if a simpler one could perform the same function. Of course, in applying this canon to ‘psychical’ functions, Morgan seems to have com- mitted himself to the view that more complex mental states require a more complex physical apparatus. 34This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? feel pain, or, to put it another way, that only human beings feel pain.20 Certainly a reasonable case could be advanced that given our admitted ignorance, we have moral grounds for giving animals the benefit of the doubt. We shall return to this point later. For the moment, let us consider the positive statement of the case. Do we have reasons for believing that only human beings feel pain? Or, recasting the question in evolutionary terms, why should pain have adaptive value for the human species, if it would serve no purpose in other species? IV Pain is a mental state, and mental states require minds. Our inquiry, then, is in part an investigation of the selective advantage conferred by the possession of a mind. A mind’s reflection on its own activities, amongst other things, enables us to predict the behaviour of other human beings, and to a lesser extent, animals. By reflecting upon our reasons for behaving in certain ways, and by assuming that our fellow human beings are similarly motivated, we can make predictions about how they are likely to behave in certain situations. But more than this, by ascribing consciousness and intelligence to other organisms we can also make predictions about how they will behave. Such ascriptions, whether they have any basis in fact or not, can thus help the human species survive. As H. S. Jennings remarked almost ninety years ago, if an amoeba ‘were as large as a whale, it is quite conceivable that occasions might arise when the attribution to it of the elemental states of consciousness might save the unsophisticated human from destruction that would result from lack of such attribution.’21 Along with human self-awareness then, came a tendency to attribute a similar awareness to other creatures. That animals might have beliefs, mental images, inten- tions and pains like our own could be nothing more than a useful fiction which gives us a shorthand method of predicting their behaviour. There is, then, some value in the belief that animals suffer pain, for it provides a reasonably reliable guide to how they will behave. But it is not an infallible guide. If, for example, we were to pit ourselves against a chess-playing computer, the best strategy to adopt would be to act as if the machine were a skilled human opponent, possessed of certain intentional states-a desire to win, particular beliefs about the rules, 20 Thus Descartes admitted in his letter to More that his thesis about animals was only probable. Philosophical Letters, 244. 21 Quoted in Larry Weisenkrantz, ‘Neurophysiology and the Nature of Consciousness’, Mindwaves, C. Blakemore and S. Greenfield (edsyf 2 [ I R U G : Blackwell, 1987yf . 35This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison and so on. However, there might be occasions when it would be better to adopt another attitude towards the computer. Let us imagine that the computer was programmed to play at three levels-beginner, inter- mediate, and advanced. Set at the ‘beginner’ level, the computer might show itself to be vulnerable to a basic ‘fool’s mate’, so that whenever this simple gambit was used, it inevitably lost. A human opponent could thus be confident of beating the computer whenever he or she wished. Now this exploitation of the computer’s weakness would result from the adoption of quite a different stance. No longer would the computer be treated as if it had desires and beliefs (or more importantly as if it had the ability to acquire new beliefsyf I R U D K X P D Q R S S R Q H Q W L Q W K H V D P e situation would quickly learn to counter the ‘fool’s mate’. Instead, predictions of the computer’s behaviour would be based on the way it had been designed to operate. Thus, our wildebeest, on an intentional account, should exhibit ‘pain behaviour’. Only when we adopt a ‘design stance’ (the animal was ‘designed’ by natural selection to behave in ways which would enhance the survival of the speciesyf G R Z H J H W D U H D V R Q D E O e explanation of why it dies in silence.22 The general point is this. The ascription to animals of certain mental states usually enables us to predict their behaviour with some accuracy (such ascription increasing our own chances of survivalyf % X W W K H U H Z L O O D O Z D V E H L Q V W D Q F H V Z K H U e this intentional model will break down and explanations which refer to selective advantages will be preferred. Another reason for attributing pain experiences only to human beings is to do with free-will and moral responsibility. While there has been some dispute about whether animals ought to be the object of our moral concern, we do not usually consider animals to be moral agents. Animals are not generally held to be morally responsible for their own acts, and notwithstanding some rather odd medieval judicial practices, animals do not stand trial for antisocial acts which they might have committed. What is absent in animals which is thought to be crucial to the committing of some wrong is the mens rea-the evil intent. Ani- mals are not morally responsible for the acts they commit because while they may have behavioural dispositions, they do not have thoughts and beliefs about what is right and wrong, nor can they, whatever their behavioural disposition, form a conscious intent. Or at least, so we generally believe. Animals, in short, are not ‘free agents’, and this is why they are not regarded as being morally responsible. But what does the determined nature of animal behaviour have to do with pain? Simply this, that if animals’ behaviours are causally determined, it makes no sense to speak of pain as an additional causal factor. 22 The terms ‘intentional stance’ and ‘design stance’ are D. C. Dennett’s. See his Brainstorms (Hassocks: Harvester Press, 1978yf . 36This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? One way of seeing the force of this is to explore some of the contexts in which we use the term ‘pain’. There are many ways we have of talking about pain which exclude animals. Consider the following: (1yf ) R U W K e long-distance runner, it is a matter of mind over matter. He must break through the pain barrier’. (2yf 7 K H K X Q J H U V W U L N H U I L Q D O O V X F F X P E H d and died’. (3yf ( Y H Q W K R X J K V K H N Q H Z L W Z R X O G P H D Q D K R U U L E O H G H D W K D t the stake, she refused to recant’. (4yf 7 K H S D L Q E H F D P H X Q E H D U D E O H + e cried out’. If we attempt to substitute animals for the human agents in these statements, the result becomes complete nonsense. Our inability to fit animals into the logic of these expressions is not merely because animals are not (contingentlyyf O R Q J G L V W D Q F H U X Q Q H U V R U K X Q J H r strikers, or religious martyrs. The key lies in statement (4yf : H P X V t ask: Do animals ever find pain unbearable?, and, What reasons could they have for bearing it? Consider this sentence in which a suitable substitution might be made. ‘The man’s hand reached into the flames, and was immediately withdrawn with a cry’. We could easily substitute ‘ape’ for ‘man’ here and the statement will retain its sense. But what about this: ‘The man plunged his hand into the flames again, knowing that only he could reach the valve and stem the flow of petrol which threatened to turn the sleepy village into an inferno.’ Now the substitution becomes impossi- ble, for what could conceivably cause the ape to plunge its hand back into the flames? Nothing, I suspect, for apes do not have reasons for bearing pain. Now it may seem unsatisfactory to proceed on the basis of certain linguistic practices to make some claim about how things really are. (This, I suspect, is why Anselm’s ontological argument always leaves one feeling a little uneasy.yf % X W W K H H [ F O X V L Y H Q D W X U H R I W K H J U D P P D U R f ‘pain’, or more correctly of ‘bearing pain’, reveals the unique province of pain. Pain operates as one kind of reason which free agents are bound to take into consideration when they decide on a particular course of action. Pain can be borne if there are reasons. But an animal never has reasons either to bear pain, or to succumb to pain. And if pain never need be brought into the sphere of reasons-the mind-then there is no need for it, qua unpleasant mental event, at all. Thus, while it is undeniable that animals sense noxious stimuli and react to them, these stimuli only need be represented as unpleasant mental states if they are to become the body’s reasons in the context of other reasons. Only as various degrees of unpleasantness can they be taken seriously amongst reasons, and this is only necessary in the mind of a rational agent. Another way of thinking about this is to consider the attributes of the long-distance runner, the hunger striker, the martyr, the hero of the sleepy village. We could say that they had mental strength, great courage, or moral character. But we would never predicate these of 37This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison animals. The wildebeest dies silently and does not endanger the herd. But does it die courageously? Does it bear the pain to the end? Does it have a reason for remaining silent? No, because it does not have a choice. All wildebeest behave in this fashion. And if it does not have a choice, there is no requirement for the dismemberment of its body to be represented mentally as pain. Pain is the body’s representative in the mind’s decision-making process. Without pain, the mind would imperil the body (as cases of insensitivity to pain clearly showyf % X W Z L W K R X W W K H U D W L R Q D O G H F L V L R Q – making mind, pain is superfluous. Animals have no rational or moral considerations which might overrule the needs of the body. It is for this reason that Descartes referred to pain, hunger and thirst as ‘confused modes of thought’, which can only be predicated of creatures which can think.23 V We may now return to the original issue which prompted this examin- ation of the reasons for ascribing pains to animals-the moral question of how we should treat animals. The arguments set out above do not constitute a conclusive disproof of animal pain. Indeed if the mind- body problem is as intractable as I have suggested, then the best we can manage is to arrive at some degree of probability. This much should be clear, however: First, there are reasons for claiming that only human beings feel pain; second, our treatment of animals cannot be based on dubious speculations about their mental lives. It follows, at the very least, that Bentham’s question cannot provide a sound basis for an ethic which is to extend to animals. How then do we proceed from here? It will seem to some that while there remains even a small possibility that animals (or certain kinds of animalsyf I H H O S D L Q W K H V H F U H D W X U H s ought to be given the benefit of the doubt. This is true to a point. Animal liberationists and animal rights activists have performed a valuable service in exposing many frivolous and mischievous practices which resulted in the unnecessary mutilation and deaths of animals. Such practices should cease, and many have. On the other hand, there are many animal experiments which improve, or might lead to the improvement of, the human lot. Even if a utilitarian equation which balances net pleasures over net pains can provide a rational basis for making moral choices in these matters (and this is doubtfulyf W K e balance should be tipped in favour of human beings, given our uncer- 23 Meditation IV (HR I, 192yf P H P S K D V L V & I 1 R U W R Q 1 H O N L Q Z K R V W D W H s that pain is an attitude not a sensation. ‘Pains and Pain Sensations’, 148. 38This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Do Animals Feel Pain? tainty about animal pain. Further, it virtually goes without saying that if it is doubtful that animals experience physical pain, even more groundless are claims that animals have other kinds of mental states- anxiety, the desire for freedom, and so on. Concerns for the psychologi- cal well-being of battery hens, veal calves, penned dolphins, and the like, would seem to be fundamentally misplaced. Our moral sen- sibilities have gone sadly awry when we expend effort on determining ‘what animals prefer’ before inquiring into whether ‘preference’ can be sensibly applied to animals. This is especially so when we are in little doubt as to what human beings prefer, and yet so many of them exist in conditions little different from those of battery hens. None of this means, however, that there are no strictures on how we ought to behave towards animals. Other considerations-aesthetic, ecological, sentimental, psychological, and pedagogical-can give us a more solid foundation for an ‘animal ethic’. Briefly, it would be morally wrong to attack Michelangelo’s ‘Pieta’ with a hammer, despite the fact that this beautifully crafted piece of marble cannot feel pain. If animals are mere machines, they are, for all that, intricate and beautiful machines (most of themyf Z K L F K O L N H R O G E X L O G L Q J V W U H H V D Q G Z R U N V R f art, can greatly enrich our lives. Accordingly, rational arguments can be mounted against acts which would damage or detroy them. There is also a growing awareness in the Western world that human beings and animals form part of a global biological community. While at times this awareness expresses itself in rather silly ways, it is still true that if we carelessly alter the balance of that community by the slaugh- ter of certain animals for pleasure or short-term economic gain, we place at risk the quality of life of ourselves and that of future generations. At a more personal level, many people form strong emotional attach- ments to animals. Domestic animals traditionally have served as play- mates for children and as company for the elderly. If mistreating these animals causes human beings to suffer, then such mistreatment is clearly wrong. Moreover, as the notorious Milgram experiments have shown, the belief that one is causing pain to another, even if false, can do great psychological harm.24 When we believe we are being cruel to animals we do ourselves damage, even though our belief might be mistaken. Finally, there is surely some value in the observation of Thomas Aquinas that kindness to animals might help to teach kindness to 24 See Stanley Milgram, Obedience to Authority: An Experimental View (London: Tavistock, 1974yf . 39This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms Peter Harrison human beings.25 Considerations of these kinds, though they require further development, can provide a far more certain guide to how we should treat animals. Bond University 25 Summa theologiae, la, 2ae. 102, 6. 40This content downloaded from 64.106.42.43 on Sun, 05 Mar 2023 20:07:25 UTC All use subject to https://about.jstor.org/terms