Moral Motivation and Moral Semantics - David O. Brink

Draft  of  10-­‐15-­‐14  

PHIL  202:  Core  Ethics   Fall  2014;  Classics  in  Metaethics   David  O.  Brink   Handout  #2:  Moral  Motivation  and  Moral  Semantics     There  is  a  puzzle  about  moral  motivation  that  we  encounter  when  we  try  to  reconcile  intellectual  and   practical   aspects   of   morality.     Cognitivists   interpret   moral   judgments   as   expressing   cognitive   attitudes,   such   as   belief.     Moral   judgments   ascribe   properties   –   axiological,   deontic,   and   aretaic   –   to   persons,   actions,  institutions,  and  policies.    Internalists  believe  that  moral  judgments  necessarily  engage  the  will   and   motivate.     We   expect   people   to   be   motivated   to   act   in   accord   with   their   moral   judgments   and   would  find  it  odd  for  people  to  be  systematically  indifferent  to  what  they  judge  morally  significant.  It  is   also   a   common   view   that   motivation   involves   pro-­‐attitudes,   such   as   desires.     We   explain   intentional   action  as  the  product  of  informational  states,  such  as  beliefs,  and  practical  states,  such  as  desires.    But   beliefs   and   desires   are   logically   independent   states;   no   belief   entails   any   particular   desire.     These   assumptions  are  in  tension.     1. 2. 3. 4.

Moral  judgments  express  beliefs.     Moral  judgments  entail  motivation.     Motivation  involves  a  desire  or  pro-­‐attitude.     There  is  no  necessary  connection  between  any  belief  and  any  desire  or  pro-­‐attitude.    

Different  views  in  metaethics  and  moral  psychology  can  be  understood  as  responses,  explicit  or  tacit,   to  the  tension  created  by  this  quartet  of  claims,  responses  that  avoid  the  tension  by  denying  at  least   one   of   the   constituent   assumptions   on   the   strength   of   others.   Noncognitivists   (e.g.   Ayer,   Stevenson,   Blackburn,   and   Gibbard)   reject   (1),   claiming   that   moral   judgments   express   desires   or   other   pro-­‐ attitudes,  rather  than  beliefs.    Externalists  (e.g.  Foot  and  Brink)  reject  (2),  claiming  that  motivation  is   not   internal   or   necessary   to   moral   judgment,   as   such.     Rationalists  about  moral  motivation   reject   (3)   or   (4),  claiming  either  that  normative  beliefs  can  motivate  without  the  benefit  of  desires  (e.g.  McDowell)   or  that  normative  beliefs  necessitate  the  desires  that  are  necessary  for  moral  motivation.  (e.g.  Nagel).         SMITH'S  MORAL  PROBLEM     Michael   Smith's   book   is   focused   on   a   related   tension,   which   he   calls   "the   moral   problem."     His   problem  consists  of  three  claims  (12).     1. Moral   judgments   of   the   form   'It   is   right   that   I   φ'   express   a   subject's   beliefs   about   an   objective   matter  of  fact,  a  fact  about  what  is  right  for  her  to  do.   2. If  someone  judges  that  it  is  right  that  she  φs  then,  ceteris  paribus,  she  is  motivated  to  φ.   3. An   agent   is   motivated   to   act   in   a   certain   way   just   in   case   she   has   an   appropriate   desire   and   a   means-­‐end  belief,  where  belief  and  desire  are,  in  Hume's  terms,  distinct  existences.     It's  worth  noting  some  differences  between  Smith's  problem  and  my  puzzle.     • Smith's  (3)  combines  the  elements  in  my  (3)  and  (4).       o These  two  elements  are  independent.    Because  rejection  of  these  two  elements  provides   two  distinct  solutions  to  the  puzzle,  it  is  useful  to  separate  them.   • Smith's  version  of  internalism  –  his  (2)  -­‐-­‐  is  weaker  than  mine  –  my  (2).   o Several   central   figures   (e.g.   Price,   Stevenson,   Harman,   Mackie,   Blackburn)   commit   themselves  to  the  stronger  internalist  thesis.   o The   trio/quartet   is   not   inconsistent   if   we   employ   the   weaker   internalist   thesis,   which   renders  arguments  that  reject  one  element  on  the  strength  of  others  invalid.    

 

2  

NONCOGNITIVISM     One   way   to   motivate   noncognitivism   (e.g.   emotivism,   prescriptivism,   expressivism)   is   as   a   solution  to  the  puzzle  about  moral  motivation  that  preserves  (2)-­‐(4).    We  can  argue  against  cognitivism   and   for   noncognitivism   by   appeal   to   internalism   and   belief-­‐desire   psychology.     There   are   various   possible  noncognitivist  views.     • Emotivism:  moral  judgments  express  (noncognitive)  feelings.   • Emotivism:  moral  judgments  express  desires  and  aversions.   • Prescriptivism:  moral  judgments  express  (universal)  prescriptions  about  how  to  feel  or  act.   • Expressivism:  moral  judgments  express  plans  for  action  (actual  or  hypothetical).     All   versions   of   noncognitivism   contrast   with   rational   intuitionism.     As   we   noted,   the   British   school   were  all  rational  intuitionists  who  combined  a  realist  and  non-­‐naturalist  metaphysics  of  morals  with   an   intuitionist   moral   epistemology.     It   was   Moore’s   OQA   that   forced   their   realism   to   take   a   non-­‐ naturalist  form.         1. If   moral   judgments   express   the   appraiser’s   beliefs   about   the   moral   properties   of   people,   actions,   and   situations,   then   these   properties   must   be   either   natural   (or   metaphysical)   properties  or  non-­‐natural  properties.   2. Moral   properties   are   not   natural   (or   metaphysical)   properties,   because   moral   predicates   are   not  definable  in  non-­‐moral  terms.    [OQA]   3. Hence,  if  moral  judgments  ascribe  properties,  they  must  be  non-­‐natural  sui  generis  properties.     Of   course,   the   rational   intuitionists   affirm   the   antecedent   in   (1)   and   (3)   and   so   defend   non-­‐naturalism.     But  one  person’s  conditional  is  another  person’s  contrapositive.    Many  noncognitivists,  including  the   emotivists,   accepted   the   OQA   and   so   accepted   (3).     But   they   found   non-­‐naturalism   a   metaphysically   and/or   epistemologically   extravagant   commitment,   and   so   they   rejected   the   antecedent   of   these   conditionals.     1. If   moral   judgments   express   the   appraiser’s   beliefs   about   the   moral   properties   of   people,   actions,   and   situations,   then   these   properties   must   be   either   natural   (or   metaphysical)   properties  or  non-­‐natural  properties.   2. Moral   properties   are   not   natural   (or   metaphysical)   properties,   because   moral   predicates   are   not  definable  in  non-­‐moral  terms.    [OQA]   3. Hence,  if  moral  judgments  ascribe  properties,  they  must  be  non-­‐natural  sui  generis  properties.   4. Non-­‐naturalism  is  false.   5. Hence,   moral   judgments   do   not   express   the   appraiser’s   beliefs   about   the   moral   properties   of   people,   actions,   and   situations.     Instead,   moral   judgments   express   noncognitive   affective   or   conative  attitudes.     So   began   the   development   of   noncognitivism.     Its   original   form   was   emotivism,   as   found   in   A.J.   Ayer   (1910-­‐89)   in   Language,  Truth,  and  Logic   (1936/46)   and   C.L   Stevenson   (1908-­‐79)   in   his   early   paper   “The   Emotive   Meaning   of   Ethical   Terms”   (1937),   in   his   book   Ethics   and   Language   (1944),   and   in   a   series  of  essays  collected  in  Facts   and   Values  (1963).    It  would  also  be  developed  as  prescriptivism  by   R.M.  Hare  (1919-­‐2002),  especially  in  The  Language  of  Morals  (1952)  and  Moral  Thinking  (1981),  and   later   as   expressivism   by   Allan   Gibbard   in   Wise  Choices,  Apt  Feelings   (1992)   and   Thinking  How  to  Live   (2003)  and  as  quasi-­‐realism  by  Simon  Blackburn  in  Ruling  Passions  (1998).     TWO  KINDS  OF  EXPRESSIVISM     There   is   an   extended   sense   of   expressivism   in   which   emotivism,   prescriptivism,   and   quasi-­‐ realism   are   all   forms   of   expressivism   insofar   as   they   claim   that   moral   judgments   fundamentally  

 

3  

express   noncognitive   emotions,   desires,   commitments,   or   plans.     But   it   is   worth   distinguishing   two   different  kinds  of  expressivism,  corresponding  roughly  to  early  and  late  expressivism.         Revisionary  expressivism  self-­‐consciously  rejects  the  realist  and  cognitivist  assumptions  of  the   moral   philosophy   of   the   British   school.     It   insists   that   despite   the   indicative   or   assertoric   syntax   and   semantics   of   moral   discourse,   moral   judgments   are   really   expressive,   rather   than   assertoric.     Moral   judgments  are  expressions  of  feeling,  and  are  not  truth-­‐apt.    There  is  no  such  thing  as  moral  truth  or   moral  knowledge.    The  most  we  can  hope  to  do  is  make  sure  that  our  moral  feelings  are  well-­‐informed   as  to  the  non-­‐moral  beliefs  on  which  they  rest.           Ayer   is   the   clearest   representative   of   this   kind   of   expressivism,   but   there   are   important   echoes   of   this   brand   of   expressivism   in   Stevenson   and   Hare.     Ayer   appeals   to   his   empiricist   theory   of   meaning   to   claim   that   cognitively   meaningful   claims   must   be   either   (a)   analytic   or   (b)   synthetic   and   empirically   verifiable.    He  concludes  that  moral  claims  are  neither;  they  are  synthetic  but  unverifiable  claims  and,   hence,  are  without  cognitive  significance.    As  a  result,  he  rejects  the  possibility  of  a  moral  science  (as   distinct  from  a  branch  of  psychology)  and  concludes  that  there  is  no  distinctive  contribution  to  ethics   that  philosophy  can  make  once  it  establishes  the  truth  of  emotivism.    Stevenson  and  Hare  draw  related   conclusions   about   the   limits   of   moral   philosophy.     Interestingly,   even   less   revisionary   expressivists,   such  as  Gibbard  and  Blackburn,  retain  the  revisionist’s  idea  that  moral  philosophy  is  the  study  of  “the   logic   of   the   moral   concepts”   (Hare’s   phrase).     At   least,   their   own   contributions   to   ethical   theory   talk   about  moral  talk.1       But  when  expressivism  takes  this  highly  revisionary  form,  it  may  seem  suspect.    For  the  syntax   and   semantic   of   moral   discourse   is   assertoric.     Moreover,   if   we   treat   moral   assertion   as   disguised   expressions   of   noncognitive   attitude,   we   encounter   problems   with   the   meaning   of   moral   phrases   in   unasserted   contexts,   such   as   the   antecedents   of   conditionals.     Further,   we   recognize   moral   truth   when   we   assert   that   some   moral   proposition   or   judgment   is   true.     We   can   perhaps   offer   expressivist   reconstructions  of  such  ascriptions  of  truth  to  moral  claims  by  adopting  a  deflationary  or  redundancy   conception   of   ascriptions   of   truth,   according   to   which   to   call   a   moral   judgment   true   is   just   to   agree   with   it.     But   it   seems   we   need   truth,   and   not   just   ascriptions   of   truth,   to   explain   valid   inferences   involving  moral  contents.    Consider  this  argument.     1. If   it   is   wrong   to   murder   innocent   children,   then   it   is   wrong   to   pay   someone   else   to   murder   innocent  children.   2. It  is  wrong  to  murder  innocent  children.   3. Hence,  it  is  wrong  to  pay  someone  else  to  murder  innocent  children.     We  can  illustrate  two  concerns  with  this  argument.         This   first   issue   concerns   the   unasserted   context   in   (1).     The   conditional   judgment   may   express   an   attitude   toward   paying   someone   to   murder   innocent   children,   although   this   attitude   would   be   conditional   on   the   wrongness   of   murdering   innocent   children.     It   expresses   no   attitude   toward   the   wrongness   of   murdering   innocent   children.     This   is   why   the   moral   claim   in   the   antecedent   of   the   conditional  in  (1)  involves  an  unasserted  context.    But  what  then  is  its  meaning?    Frege  believed  that   an  adequate  semantic  theory  should  recognize  the  univocity  of  terms  across  asserted  and  unasserted   contexts   of   utterance.     Peter   Geach   thought   that   Frege’s   point   posed   problems   for   noncognitivist   analyses   of   the   meaning   of   moral   judgments.2     The   expressivist   construes   moral   assertion   as   the   expression  of  the  appraiser’s  attitudes,  rather  than  a  description  of  the  way  the  world  is.    But  then  it  is   not   clear   how   the   noncognitivist   understands   the   meaning   of   moral   predicates   in   unasserted   contexts,   such   as   the   antecedents   of   conditional   statements.     And   this   is   relevant   to   the   validity   of   the   argument.     For  the  argument  to  be  valid,  the  relevant  moral  claims  must  have  the  same  meaning  in  asserted  and   unasserted   contexts.     It’s   hard   to   see   how   this   could   be   true   if   we   accept   the   expressivist   analysis   of  

                                                                                                               

1  It  is  perhaps  significant  that  Gibbard’s  second  book  is  entitled  Thinking  How  to  Live,  rather  than  How  to  Live.   2  Peter  Geach,  “Assertion”  Philosophical  Review  74  (1965):  449-­‐65.  

 

4  

moral   assertion.     It’s   easy   to   explain   if   moral   propositions   have   the   same   indicative   meaning   in   asserted  and  unasserted  contexts.       The  second  issue  is  that  to  explain  the  validity  of  the  argument  we  apparently  need  to  invoke   truth.     The   reason   the   argument   is   valid   is   that   it   is   not   possible   for   the   premises   to   be   true   and   the   conclusion  false.    This  explanation  of  validity  is  truth-­‐theoretic.    Here,  it  seems,  we  need  truth  and  not   just  a  deflationary  semantics  for  ascriptions  of  truth.     This   is   a   condensed   summary   of   some   standard   worries   about   the   semantic   and   logical   resources   of   revisionary   expressivism.     The   conditional/contrapositive   gambit   can   be   recycled.     If   expressivism   leads   to   all   these   semantic   and   logical   problems,   then   that   might   be   reason   to   reject   expressivism   and   accept   cognitivism   and   realism   after   all.     Some   might   want   to   resurrect   non-­‐ naturalism.    Others  will  think  that  this  makes  a  strong  case  for  ethical  naturalism,  especially  since  we   should   reject   the   OQA   in   any   case,   in   particular,   rejecting   the   semantic   test   of   properties,   the   assumption  that  meaning  should  be  transparent,  or  both.     But,   of   course,   the   expressivist   might   think   that   these   semantic   and   logical   problems   for   revisionary  expressivism  are  not  decisive.    These  problems  motivate  a  different  kind  of  expressivism.     Accommodationist  expressivism  tries  to  show  how  there  can  be  semantic  and  logical  relations  among   our  noncognitive  attitudes  toward  moral  contents,  so  that  we  can  reconstruct  or  mimic  validity,  truth,   and  knowledge  in  moral  matters  within  an  expressivist  framework.         Revisionary   and   accommodationist   expressivism   are   poles   in   a   spectrum.     Though   there   are   accommodationist   elements   in   both   Stevenson   and   Hare,   I   think   their   form   of   expressivism   is   still   predominantly   revisionary.     Expressivism   has   assumed   a   robustly   accommodationist   character   only   in   the  work  of  Gibbard  and  Blackburn.     This   is   not   the   place   to   explore   or   assess   the   adequacy   of   accommodationist   expressivism,   which   is   an   extremely   complicated   matter.3     But   I   thought   it   might   be   useful   to   situate   the   more   revisionary  form  of  expressivism  contained  in  the  emotivism  of  Ayer  and  Stevenson  within  this  larger   dialectical  picture.     AYER     In   Language,   Truth,   and   Logic   Ayer   applies   his   empiricist   theory   of   meaning   to   moral   judgments   and   concludes   that   they   lack   cognitive   significance,   because   they   are   neither   analytic   nor   empirically   verifiable.     They   are   not   analytic,   as   shown   by   a   variation   of   the   OQA   (104-­‐05).     Whereas   it   is  “self-­‐contradictory”  to  deny  that  M-­‐things  are  M,  it  is  never  self-­‐contradictory  to  deny  that  N-­‐things   are  M.    Ayer  assumes  that  moral  claims  are  not  empirically  testable,  apparently  without  argument.     We  might  dispute  this.    We  might  think  that  moral  claims  are  empirically  testable,  but  only  in   conjunction  with  suitable  auxiliary  hypotheses.     1. Good  people  keep  their  promises  even  when  this  requires  modest  personal  sacrifice.   2. Ben  made  a  promise  to  Sam  that  can  only  be  kept  at  modest  personal  sacrifice.   3. Ben  is  a  good  person.   4. Hence,  Ben  will  keep  his  promise  to  Sam.     Moral  claims  have  no  observational  consequences  by  themselves  but  only  in  conjunction  with  auxiliary   hypotheses,  which  include  moral  claims  for  which  we  can  have  independent  evidence.    The  situation   here   is   no   different   than   it   is   in   non-­‐moral   matters.     If   Ben   does   not   keep   his   promise   to   Sam,   then   something  will  have  to  go   -­‐-­‐  I   will   need   to   revisit   my   assumptions   about   the   promise-­‐keeping   behavior   of   good   people,   my   assumptions   about   Ben’s   character,   or   perhaps   my   assumptions   about   the   sacrifice   required  of  Ben.    This  is  just  a  familiar  feature  of  holism  about  confirmation.  

                                                                                                                3  See  

my   review   of   Gibbard’s   Thinking   How   to   Live   in   Philosophical   Review   116   (2007):   267-­‐72   and   Mark   Schroeder,  Being  For:  Evaluating  the  Semantic  Program  of  Expressivism  (Oxford:  Clarendon  Press,  2008).  

 

5  

  Moreover,   some   ethical   naturalists   have   wanted   to   resist   Harman’s   suggestion   that   moral   claims   are   explanatorily   impotent.4     If   moral   facts   had   explanatory   value,   this   would   apparently   also   cast  doubt  on  Ayer’s  assumption  that  moral  claims  are  not  empirically  testable.       If   moral   judgments   are   cognitively   meaningless,   Ayer   concludes,   they   must   have   some   other   kind  of  meaning.    They  must  have  a  purely  emotive  meaning,  expressing  the  appraiser’s  emotions.    As   such,  moral  judgments  would  be  neither  true  nor  false  (103,  107,  108).    In  addition  to  expressing  the   appraiser’s  noncognitive  feelings,  Ayer  claims  that  moral  judgments  can  aim  and  succeed  at  arousing   similar  feelings  in  others  (108).         Ayer   contrasts   his   emotivist   metaethics,   according   to   which   moral   judgments   express   the   appraiser’s   attitudes,   with   a   subjectivist   metaethics,   according   to   which   moral   judgments   report   the   appraiser’s  attitudes.    According  to  subjectivism,  when  I  say     Torturing  babies  for  fun  is  wrong.     I  am  reporting  that  I  have  a  negative  feeling  toward  the  conduct  in  question.    In  effect,  I  am  saying     I  don’t  like  the  torturing  of  babies  for  fun.     Or  perhaps     I  disapprove  of  torturing  babies  for  fun.     By  contrast,  according  to  emotivism  (and,  more  generally,  expressivism),  I  am  not  reporting  that  I  have   this   attitude,   but   rather   expressing   it.     For   example,   I   often   do   things   that   express   my   anger,   such   as   swearing   or   grimacing,   without   reporting   that   I   am   angry,   though,   in   normal   circumstances,   you   might   well  be  able  to  infer  that  I  am  angry  from  my  expression  of  anger.    On  this  view,  I  am  expressing  an   attitude  of  disapproval  toward  the  conduct  in  question.    In  effect,  I  am  saying       Booh!  (torturing  babies  for  fun)     If   we   look   for   ordinary   linguistic   expressions   of   these   attitudes,   we   might   look   to   the   imperatival   form   (103).     Don’t  torture  babies  just  for  fun!       Ayer   notes   that   the   subjectivist   cannot   represent   moral   disagreement.     Consider   a   disagreement   between  Axel  and  Bert.    Suppose  Axel  affirms  and  Bert  denies  our  moral  proposition.     Torturing  babies  for  fun  is  wrong.     The   subjectivist   believes   that   Axel   and   Bert   is   each   reporting   his   attitude   toward   the   conduct   in   question.    Axel  is  saying  that  he  disapproves,  and  Bert  is  denying  that  he  disapproves.    But  there  is  no   disagreement  here.    They  disagree  no  more  than  if  Axel  said  he  likes  sushi  and  Bert  said  he  detests  it.     Disagreement  seems  to  require  asserting  inconsistent  things  -­‐-­‐  things  that  can’t  both  be  true.    But  what   Axel  and  Bert  assert,  according  to  the  subjectivist,  are  not  inconsistent.     Stevenson   will   claim   that   the   emotivist   has   a   better   account   of   moral   disagreement   than   the   subjectivist.     But   Ayer   thinks   that   the   emotivist   should   agree   with   subjectivist   that   genuine   moral   disagreement   is   not   possible   (110-­‐12).     Ayer   thinks   that   there   can   be   moral   disputes   that   reduce   to  

                                                                                                               

4  See  Gilbert  Harman,  The  Nature  of  Morality  (OUP,  1977),  chs.  1-­‐2.  

 

6  

disagreements  about  the  non-­‐moral  facts  on  which  our  attitudes  depend  and  that  we  can  argue  with   others   about   moral   matters   insofar   as   we   can   provide   evidence   for   non-­‐moral   claims   on   which   our   moral   attitudes   depend.     Thus,   we   can   make   sense   of   moral   reasoning   within   a   shared   scheme   of   attitudes   as   reasoning   about   the   non-­‐moral   facts   relevant   to   applying   our   shared   attitudes.     But   it   is   not   possible,   Ayer   thinks,   to   reason   or   disagree   about   fundamental   attitudes.     These   are   not   genuine   disputes.     In  making  these  claims,  Ayer  ignores  the  realist  idea  that  we  can  be  mistaken  not  only  about   derivative   moral   commitments,   as   the   result   of   mistaken   non-­‐moral   assumptions   with   which   we   apply   fundamental   evaluative   commitments,   but   also   about   fundamental   evaluative   commitments.     He   also   ignores   the   idea   that   we   can   reason   about   our   fundamental   evaluative   commitments   in   a   dialectical   fashion.     STEVENSON         Stevenson   develops   his   conception   of   emotivism   in   his   early   paper   “The   Emotive   Meaning   of   Ethical  Terms,”  in  his  book  Ethics  and  Language,  and  in  a  series  of  essay  collected  in  Facts  and  Values.     We  will  concentrate  on  general  features  of  Stevenson’s  views.     Stevenson  sets  out  three  desiderata  for  any  conception  of  the  meaning  of  moral  terms  and  the   function  of  moral  judgments  (EMET  16-­‐17).     1. It  must  make  sense  of  moral  disagreement.   2. It  must  make  sense  of  the  “magnetic”  or  “dynamic”  aspects  of  moral  judgments,  their  relation   to  motivation  or  the  will  and  their  role  in  interpersonal  engagement.   3. It  must  vindicate  the  autonomy  of  ethics  in  relation  to  the  sciences.     Like  Ayer,  Stevenson  contrasts  emotivism  with  subjectivism  or  what  he  sometimes  calls  “the  interest   theory”  (15).    Subjectivism  fares  badly,  Stevenson  thinks,  on  all  three  grounds.     Emotivism,   Stevenson   claims,   understands   moral   judgments   primarily   to   be   expressing,   rather   than   reporting,   the   appraiser’s   noncognitive   emotions   and   desires   and   attempting   to   arouse   similar   attitudes   in   the   appraiser’s   audience   (18-­‐19).     Notice   that   there   are   two   distinguishable   functions   alluded   to   here:   expressing   one’s   own   attitudes   and   arousing   similar   attitudes   in   others   or,   as   Stevenson  sometimes  says,  influencing  others.    One  could  apparently  aim  for  either  goal  without  the   other.     Though   Stevenson   sometimes   suggests   that   the   interpersonal   aim   is   primary   (19),   it’s   not   clear   why  either  wouldn’t  do.    If  so,  we  might  represent  the  emotivist  claim  about  the  primary  function  of   moral  judgments  disjunctively:  to  express  the  appraiser’s  own  attitudes  or  to  influence  the  attitudes  of   others.    In  “The  Emotive  Meaning  of  Ethical  Terms”  Stevenson  suggests  that  subjectivism  is  incomplete,   leaving   out   the   primary   emotive   meaning   of   ethical   terms.     On   one   interpretation,   emotivism   supplements   subjectivism   with   a   distinctively   emotive   element.     This   conception   of   emotivism   is   roughly  what  Stevenson  will  later  identify  as  the  First  Pattern  of  Analysis:       To  say  “X  is  good”  is  roughly  synonymous  with  “I  approve  of  X;  do  so  as  well”  [EL  81].     So  understood,  emotivism  is  superior  to  simple  subjectivism  along  all  three  dimensions.     Like   Ayer,   Stevenson   recognizes   that   subjectivism   cannot   represent   moral   disagreement.     If   moral   judgments   are   just   reports   of   the   appraiser’s   attitudes,   then   it’s   not   clear   why   Axel   and   Bert   disagree;  each  is  reporting  his  own  attitudes  toward  the  conduct  in  question.    Unlike  Ayer,  Stevenson   takes  this  to  be  a  problem  for  subjectivism.    Also  unlike  Ayer,  Stevenson  thinks  that  the  emotivist  can   and  should  have  an  account  of  moral  disagreement.    The  emotivist  understands  moral  disagreement,   not   as   the   assertion   of   incompatible   propositions,   but   as   the   expression   of   conflicting   attitudes   (19-­‐20,   26-­‐27).    According  to  the  emotivist,  Axel  and  Bert  are  not  asserting  anything,  much  less  inconsistent   things.    They  are  merely  expressing  contrary  attitudes.    Their  disagreement  is  like  the  one  at  a  sushi   bar  when  two  friends  look  at  the  offerings  and  one  says  “Yum!”  and  the  other  says  “Gross!”.    Stevenson  

 

7  

also  says  the  disagreement  might  be  understood  as  like  the  practical  disagreement  between  Axel  and   Bert  if  Axel  wants  to  go  to  the  beach  and  Bert  wants  to  go  to  the  mountains  for  a  hike.         Subjectivism  also  seems  ill-­‐suited  to  explain  the  dynamic  aspects  of  moral  judgments,  because   it’s  not  clear  why  your  liking  something  is  a  reason  for  me  to  like  something.    But  if  moral  judgments   not  only  report  the  appraiser’s  attitudes  but  also  instruct  others  to  join  in,  then  we  can  see  how  they   have  a  dynamic  interpersonal  function  (27).     Subjectivism   also   fails   to   represent   the   autonomy   of   ethics   insofar   as   it   implies   that   an   appraiser   can   support   her   moral   judgment   simply   by   pointing   out   the   psychological   fact   that   she   approves  of  the  conduct  in  question.    But  that  is  not  how  we  try  to  support  moral  judgments.    We  don’t   cite   autobiographical   facts   about   ourselves   to   explain   why   apartheid   is   unjust.     By   contrast,   emotivism   maintains   the   autonomy   of   ethics.     While   reasoning   about   the   non-­‐moral   facts   may   help   resolve   disagreements  when  we  share  fundamental  attitudes  and  disagree  only  about  what  further  attitudes   the  non-­‐moral  facts  recommend,  not  all  disagreements  are  disagreements  about  the  non-­‐moral  facts.     There   are   also   disagreements   in   fundamental   attitudes,   and   these   disagreements   won’t   be   resolved   by   the  evidence  of  the  sciences  (27-­‐29).           According  to  Stevenson,  moral  terms  have  both  descriptive  meaning  and  emotive  meaning,  but   their   emotive   meaning   is   primary.     This   is   reflected   in   the   First   Pattern   of   Analysis.   But   the   descriptive   element  of  the  FPA  is  the  autobiographical  element  that  it  shares  with  the  subjectivist  analysis.    But  the   subjectivist   analysis   goes   wrong   by   making   moral   judgments   autobiographical.     When   I   judge   apartheid  to  be  unjust,  I  don’t  seem  to  be  making  any  claim  about  myself  or  my  attitudes.    I  seem  to  be   making   a   claim   about   apartheid   on   which   my   attitudes   are   perhaps   consequential.     Considerations   such  as  these,  led  Stevenson  to  introduce  in  Ethics  and  Language  a  Second  Pattern  of  Analysis.           “This  is  good”  has  the  meaning  of  “This  has  qualities  or  relations  X,  Y,  Z  …,”  except  that  “good”   has   as   well   a   laudatory   emotive   meaning   which   permits   it   to   express   the   speaker’s   approval,   and  tends  to  evoke  the  approval  of  the  hearer  [207].     Adapting  ideas  from  Hare,  we  might  interpret  Stevenson’s  SPA  as  distinguishing  descriptive  criteria  for   emotive  meaning  and  emotive  meaning.    On  this  view,  any  moral  judgment  picks  out  certain  features   of   a   person,   action,   or   situation   in   virtue   of   which   the   appraiser   expresses   her   attitudes   (e.g.   approval)   toward  that  person,  action  or  situation  and/or  encourages  others  to  adopt  those  attitudes.    Notice  that   SPA   drops   the   subjectivist   element   in   FPA.     Emotivism,   according   to   SPA,   is   not   a   supplement   to   subjectivism,  however  important.     QUESTIONS     Notice  that  emotivism  analyzes  moral  judgments  into  the  attitudes  of  appraisers.    The  standard   analysans   of   emotivism   are   approval/disapproval.     But   there   are   a   great   many   apparently   distinct   kinds   of   moral   judgments   possible:   judging   good,   judging   right,   judging   virtuous,   judging   admirable,   etc.     It   would   seem   that   emotivism   is   committed   to   pairing   distinct   moral   judgments   with   distinct   attitudes.     It’s   not   clear   that   we   can   identify   sufficiently   many   distinct   attitudes   for   this   purpose   without   invoking   the   very   moral   properties   to   individuate   attitudes   that   we   are   trying   to   analyze.     I   don’t  say  that  this  is  an  insurmountable  challenge  for  expressivism,  but  it  is  a  challenge.     Emotivists  seem  to  subscribe  to  internalism  about  moral  motivation,  claiming  that  motivational   states,   such   as   pro-­‐attitudes,   are   essential   parts   of   moral   judgments.     But   this   seems   to   render   impossible  disengaged   moral   judgments.    We  often  represent  the  possibility  of  amoralist  skepticism  -­‐-­‐   skepticism   about   the   authority   of   moral   requirements   -­‐-­‐   by   the   amoralist   -­‐-­‐   someone   who   professes   indifference  to  the  moral  requirements  that  she  recognizes.    This  is  a  philosophical  amoralist,  and  we   might   think   that   it   is   a   problem   with   emotivism   or   any   other   form   of   internalism   that   denies   this   possibility.    Moreover,  we  might  want  to  recognize  non-­‐philosophical  amoralists  who  are  indifferent  to   what  they  judge  morally  required,  not  out  of  philosophical  principle,  but  out  of  apathy,  depression,  or  

 

8  

the   sort   of   volitional   impairment   that   we   find   in   those   with   damage   to   the   prefrontal   cortex   of   the   brain  (e.g.  Phineas  Gage).     Stevenson  thinks  emotivism  superior  to  subjectivism,  because  it  has  a  better  account  of  moral   disagreement.     We   might   concede   this   comparative   claim,   but   wonder   if   the   emotivist   account   of   disagreement   is   as   good   as   the   realist   account.     The   realist   interprets   Axel   as   ascribing   the   property   of   impermissibility  to  the  act  type  .    By  contrast,  Bert  denies  this  claim.    They   assert   inconsistent   propositions,   which   can’t   both   be   true.     By   contrast,   the   emotivist   sees   this   as   a   conflict   of   attitudes,   like   the   conflict   between   wanting   to   go   to   the   beach   and   wanting   to   go   to   the   mountains.    It’s  not  clear  if  this  practical  conflict  is  all  that’s  involved  in  a  moral  disagreement.    Both   impulses  can  be  understandable,  even  reasonable.    By  contrast,  it’s  hard  to  believe  that  both  Axel’s  and   Bert’s  claims  about  torturing  babies  for  fun  can  be  understandable  and  reasonable.    It  seems  that  one   of  them  must  be  right  and  the  other  wrong.     The  emotivist  must  also  count  moral  disagreements  in  unusual  ways.    The  emotivist  identifies   moral  agreement  with  agreement  in  noncognitive  attitude  and  moral  disagreement  with  disagreement   in   noncognitive   attitude.     Consider   the   moralist   and   the   amoralist   who   intuitively   share   the   same   moral   views.     God   believes   that   compliance   with   the   Decalogue   is   right   and   approves   those   who   are   compliant.    Satan  also  thinks  that  compliance  with  the  Decalogue  is  morally  right  and  disapproves  of   those  who  are  compliant  for  that  very  reason.    The  emotivist  must  say  that  God  and  Satan  have  a  moral   disagreement,   because   they   have   conflicting   attitudes   toward   the   same   actions.     But   it   seems   that   they   agree  about  what  morality  requires  and  disagree  only  in  attitude.     There   are   also   the   worries   about   revisionary   expressivism   and   the   problems   it   has   with   the   semantics   and   logic   of   moral   assertion   and   inference.     Most   of   these   problems   would   seem   to   arise   for   Stevenson   since   he   treats   moral   judgments,   for   the   most   part,   as   being   not   truth-­‐apt.     In   Facts  &  Values   Stevenson   does   help   himself   to   a   deflationary   or   disquotational   account   of   ascriptions   of   truth   (216-­‐ 19).    But  while  this  may  allow  him  to  make  emotivist  sense  of  ascriptions  of  truth  to  moral  judgments   made  by  oneself  or  others,  it  is  not  clear  that  this  entitles  him  to  the  property  of  truth,  necessary,  for   instance,   for   the   truth-­‐theoretic   explanation   of   validity.     I   don’t   mean   to   suggest   that   there   are   no   emotivist   resources   to   reconstruct   the   relevant   semantic   and   logical   structures   and   relations.     But   this   is  the  project  of  accommodationist  expressivism,  pursued  by  later  expressivists,  such  as  Gibbard  and   Blackburn,  not  by  Stevenson  himself.       RATIONALIST  MORAL  MOTIVATION     We  have  looked  at  worries  about  noncognitivism.    The  externalist  solution  may  also  seem  to  be  a   solution  of  last  resort,  because  it  may  seem  to  deny  the  platitude  that  moral  judgments  are  motivation-­‐ ally   efficacious.     For   this   reason,   we   might   look   seriously   at   rationalist   theories   of   moral   motivation,   because   they   promise   to   represent   moral   judgments   as   intrinsically   motivational   without   giving   up   cognitivism.       Some   philosophers,   such   as   Nagel   and   McDowell,   maintain   cognitivism   and   internalism   about   motivation  by  rejecting  the  assumption  that  motivation  requires  a  desire  or  pro-­‐attitude;  they  insist  that   purely  cognitive  states  -­‐-­‐  beliefs  -­‐-­‐  can  motivate.    For  instance,  both  Nagel  and  McDowell  motivate  this   rationalist   view   about   moral   motivation   by   appeal   to   an   analogy   between   prudential   and   moral   motivation.     Seeing   what's   in   one's   interest   or   what   morality   requires   is   often   all   that's   needed   to   motivate   and   explain   action.     In   such   cases,   ascription   of   the   desire   to   perform   the   act   in   question   is   "merely  consequential."     However,  even  if  ascription  of  proximate  desires  is  consequential,  motivation  seems  to  depend   on   more   ultimate   desires.     Proximate   or   foreground   belief   can   be   sufficient   for   motivation   only   when   more  ultimate  desires  are  part  of  the  background.    If  we  count  appeal  to  my  belief  that  it  is  raining  as   explaining  my  action,  this  is  only  because  we  take  my  desires  to  go  out  and  to  stay  dry  for  granted  as  part   of  my  psychological  background.    Likewise,  we  can  agree  that  moral  beliefs  are  sufficient  to  motivate  the   virtuous   person,   but   this   is   only   because   a   virtuous   person   is   someone   with   a   certain   well-­‐developed   psychological  profile  that  is  structured  by  various  cognitive  and  conative  states.  

 

9  

  In  support  of  the  need  for  prior  conative  structure  in  order  for  normative  belief  to  motivate,  we   might   make   several   observations.     First,   this   picture   is   supported   by   the   fact   that   acquiring   the   same   belief  about  what  morality  requires  can  produce  different  behavior  and  motivation  in  the  virtuous  and   the   non-­‐virtuous.     Same   input,   different   output.     By   itself,   this   just   means   that   there   must   be   some   difference   in   internal   psychological   architecture   between   the   virtuous   and   the   non-­‐virtuous.     The   differences   might   be   cognitive,   not   conative.     However,   we   can   presumably   find   cases   of   differential   output   (motivation)   with   common   input   (moral   beliefs),   despite   common   beliefs.     Compare   the   virtuous   person,  who  tracks  the  morally  good,  with  the  vicious  person,  who  tracks  the  morally  bad.    Same  input,   same   cognitive   architecture,   different   output.     If   so,   then   there   must   be   a   difference   in   conative   architecture.         The   implausibility   of   the   purely   cognitive   picture   of   moral   motivation   is   a   consequence   of   the   plausibility   of   belief-­‐desire   psychology.     Proponents   of   belief-­‐desire   psychology   often   characterize   the   difference  between  beliefs  and  desires  in  terms  of  their  different  directions  of  fit  with  the  world  (e.g.   Aristotle,   Green,   Anscombe,   Stampe,   Humberstone,   Velleman).     The   difference   between   beliefs   and   desires  is  a  special  case  of  a  more  general  difference  between  representations  and  pro-­‐attitudes.    On   this  view,  representations,  such  as  beliefs,  are  states  of  the  agent  whose  content  she  adjusts  to  conform   to   information   she   receives   about   the   state   of   the   world.     By   contrast,   pro-­‐attitudes,   such   as   desires,   are  states  of  the  agent  on  the  basis  of  which  she  acts  to  make  the  world  conform  to  them.    We  can  think   of   the   difference   in   terms   of   the   response   to   a   perceived   mismatch   between   the   content   of   the   intentional   state   and   information   about   the   way   the   world   is.     If   the   state   is   a   belief,   the   agent   tends   to   respond   to   such   a   mismatch   by   modifying   the   content   of   the   intentional   state   to   match   the   way   the   world  is  or  appears.    If  the  state  is  a  desire,  the  agent  tends  to  respond  to  such  a  mismatch  by  acting  so   as   to   modify   the   world   to   conform   to   the   content   of   the   state.     On   this   sort   of   belief-­‐desire  psychology,   agents   act   in   order   to   satisfy   their   desires   based   on   their   beliefs   about   the   world,   in   particular,   their   beliefs  about  the  causal  means  to  and  necessary  conditions  of  satisfying  their  desires.     MORAL  MOTIVATION  BY  RATIONALIZED  DESIRE     A  different  form  of  rationalism  denies  that  motivation  is  possible  without  a  prior  pro-­‐attitude  but   insists  that  certain  beliefs,  in  particular,  normative  beliefs  entail  pro-­‐attitudes.    On  this  view,  the  belief   that   I   have   a   moral   reason   generates   a   desire   to   perform   the   action   in   question.     This   solution   to   the   puzzle  has  two  component  parts.     • Normative  Internalism:  beliefs  about  one's  practical  reason  entail  (pro  tanto)  desire.   • Conceptual  Rationalism:  belief  that  an  action  is  morally  required  entails  belief  that  it  is  rationally   authoritative.     Nagel  and  Smith  both  seem  to  endorse  this  sort  of  moral  motivation  by  rational  desire.    Smith's  version:       • Normative  Internalism:  To  judge  that  I  have  reason  to  φ  is  to  judge  that  I  would  desire  to  φ  if  I   was   fully   rational,   that   is,   deliberating   correctly   (e.g.   with   full   information,   coherently,   and   employing   means-­‐ends   reasoning)   (155-­‐61).     (This   is,   I   think,   a   conceptual   claim   about   normative  judgment,  not  just  a  claim  about  the  truth-­‐conditions  of  normative  judgment.)   • Conceptual  Rationalism:  To  judge  that  I  am  morally  required  to  φ  is  to  judge  that  I  would  desire   to  φ  if  I  was  fully  rational  and  that  φ-­‐ing  "is  an  act  of  the  appropriate  substantive  kind"  (184).     (This  is  also  a  conceptual  claim.)       However,  I'm  not  sure  the  details  of  Smith's  version  of  motivation  by  rationalized  desire  need  concern  us.     Though  there  is  an  important  grain  of  truth  in  normative  internalism,  both  theses  are  open  to  question.     Normative   internalism   implies   that   there   is   an   important   connection   between   judgments   of   practical  reason  and  the  will.    This  idea  might  be  defended  by  appeal  to  belief-­‐desire  psychology.    On  this   view,  as  we  have  seen,  intentional  action  is  viewed  as  the  product  of  representational  states,  such  as  

 

10  

belief,  and  pro-­‐attitudes,  such  as  desire,  which  display  different  directions  of  fit  with  the  world.  On  this   sort  of  belief-­‐desire  psychology,  agents  act  in  order  to  satisfy  their  desires  based  on  their  beliefs  about   the  world,  in  particular,  their  beliefs  about  the  causal  means  to  and  necessary  conditions  of  satisfying   their   desires.     But   on   this   sort   of   psychology,   we   can   also   understand   how   normative   beliefs   would   tend   to   influence   desire.     For   normative   beliefs   are   beliefs   about   how   the   world   should   be.     But   if   desires   are   precisely   states   that   tend   to   make   agents   modify   the   world   in   accordance   with   their   content,   then   we   should   expect   normative   beliefs   normally   to   affect   desires.     If   we   accept   belief-­‐desire   psychology,  because  of  their  different  directions  of  fit,  we  can  claim  that  desire  can  be  responsive  to   ought  judgments.      This  shows  how  one  can  accept  the  Humean  dictum  that  action  depends  on  desire   without  accepting  the  Humean  dictum  that  reason  can  only  be  the  slave  of  the  passions.     However,  to  say  that  normative  beliefs  can  and  normally  do  influence  desire  is  not  to  say  that   normative   beliefs   have   such   influence   necessarily.     Other   things   being   equal,   normative   beliefs   have   conative  influence.    But  other  things  need  not  be  equal  if  there  is  some  relevant  form  of  psychological   interference.     In   some   cases   of   weakness   of   will,   normative   beliefs   apparently   motivate   but   provide   insufficient  motivation.    In  other  cases  of  weakness  of  will,  normative  beliefs  may  not  motivate  at  all.     This   second   sort   of   weakness   of   will   is   selective   if   the   interference   is   intermittent;   it   is   global   if   the   interference   is   systematic.     Depression   might   produce   selective   weakness   of   will   (cf.   Stocker,   Mele),   but  damage  to  the  prefontal  lobe  of  the  cerebral  cortex  (as  in  the  famous  case  of  Phineas  Gage)  might   produce  systematic  weakness  of  will.       This   is   already   to   recognize   one   kind   of   amoralism   -­‐-­‐   someone   who   recognizes   moral   requirements  yet  remains  indifferent.    The  kind  of  amoralism  whose  possibility  Smith  concedes  is  a  kind   of  unprincipled  amoralism.    It  is  unprincipled,  not  in  the  sense  that  it  is  random  or  lacks  a  psychological   explanation,   but   in   the   sense   that   it   is   due   to   psychological   interference   with   the   normal   process   by   which   results   of   practical   deliberation   affect   an   agent's   motivational   set;   indifference   does   not   reflect   principles  the  agent  accepts.     The   possibility   of   this   sort   of   amoralism   undermines   the   strong   internalist   assumption   about   moral   motivation.     If,   as   I   believe,   the   puzzle   about   moral   motivation   is   best   construed   as   resting   on   this   strong   internalist   assumption,   then   recognition   of   the   possibility   of   this   sort   of   amoralism   is   sufficient   to   vindicate  the  externalist  solution.             But   there   is   another   problem   for   this   form   of   rationalism   about   moral   motivation,   even   if   we   accepted  normative  internalism.    Conceptual  rationalism  seems  problematic.    Doubts  about  conceptual   rationalism   reflect   the   possibility   of   principled   amoralism   in   which   one   is   indifferent   to   moral   requirements,  either  globally  or  selectively,  because  one  has  doubts  about  the  rational  authority  of  moral   requirements.    In  this  way  doubts  about  conceptual  rationalism  actually  presuppose  (weak)  normative   internalism.    It  is  possible  to  remain  unmoved  by  what  one  judges  morally  required  as  long  as  one  can   hold   conceptions   of   morality   and   practical   reason   according   to   which   moral   requirements,   globally   or   selectively,  lack  rational  authority.     There   are   many   conceptions   of   morality   and   practical   reason   according   to   which   their   coincidence  is  at  best  imperfect.    One  familiar  set  of  conceptions  involves  recognizing  impartial  or  other-­‐ regarding   duties   and   an   instrumental   or   prudential   conception   of   practical   reason.     For   even   if   it   is   often   true   that   fulfilling   other-­‐regarding   duties   would   satisfy   one's   desires   or   promote   one's   interests,   this   coincidence  need  not  always  take  place.    Fulfilling  one's  other-­‐regarding  duties  need  not  always  satisfy   one's  desires  or  promote  one's  interests.     Notice  that  principled  amoralism  does  not  require  the  truth  of  anti-­‐rationalism.    It  requires  only   the   denial   of   conceptual   rationalism   and   the   possibility   that   one   be   able   to   form   conceptions   of   morality   and  practical  reason  according  to  which  their  demands  can  diverge.     EXTERNALISM  ABOUT  MORAL  MOTIVATION     Doubts   about   the   plausibility   of   moral   motivation   by   rationalized   desire   tend   also   to   support   externalism  about  moral  motivation.    

 

11   • •

Unprincipled  Amoralism  undermines  strong  internalism  about  moral  motivation.   Principled  Amoralism  undermines  both  strong  and  weak  internalism  about  moral  motivation.  

  As  long  as  anti-­‐rationalist  beliefs  are  possible,  whether  or  not  they  are  true,  internalism  about  moral   motivation  must  be  false.     But   we   might   doubt   the   coherence   of   externalism,   since   it   depends   on   the   coherence   of   anti-­‐ rationalism.    Consider  this  worry  about  anti-­‐rationalism.     1. If  I  am  under  a  moral  requirement  to  φ,  there  is  a  moral  reason  for  me  to  φ.   2. If  there  is  a  moral  reason  for  me  to  φ,  there  is  a  reason  for  me  to  φ.   3. If  there  is  a  reason  for  me  to  φ,  it  would  be  pro  tanto  irrational  for  me  to  fail  to  φ.   4. Hence,  if  I  am  under  a  moral  requirement  to  φ,  it  would  be  pro  tanto  irrational  for  me  to  fail  to  φ.     This   argument   would   show   weak   internalism   to   be   a   conceptual   truth.     But   it   is   not   compelling.     Sometimes  when  we  say  that  I  have  a  reason  to  φ,  we  mean       • (a)  There  is  a  behavioral  norm  that  enjoins  φ-­‐ing  and  applies  to  me.     In   this   sense   of   reason,   moral   norms   do   imply   reasons.     There   are   as   many   kinds   of   reasons   as   there   are   norms,  including  moral  reasons,  legal  reasons,  reasons  of  etiquette.    But  we  often  have  something  more   in  mind  in  ascribing  reasons.     • (b)   There   is   a   behavioral   norm   that   enjoins   φ-­‐ing,   it   applies   to   me,   and   it   would   be   pro  tanto   irrational  for  me  not  to  φ.       If  there  is  reason,  in  this  sense,  to  act  on  a  norm,  then  practical  reason  endorses  this  norm.    But  not  all   reasons  for  action  in  the  first  sense  are  reasons  for  action  in  the  second  sense  (cf.  Foot,  "Morality  as  a   System   of   Hypothetical   Imperatives").     For   instance,   it   is   arguable   that   failure   to   conform   to   requirements   or   reasons   of   etiquette   or   law   need   not   be   pro   tanto   irrational.     It   is   clear   that   moral   requirements  are  moral  reasons  and  that  moral  reasons  are  (a)-­‐reasons.    It  is  not  clear  that  they  are  (b)-­‐ reasons.  So  the  sense  of  reason  for  action  in  which  (3)  is  true  need  not  be  the  same  sense  of  reason  for   action  in  which  (2)  is  true.    If  so,  it's  arguable  that  it  is  only  by  failing  to  distinguish  these  two  senses  of   reason  for  action  that  the  rational  authority  of  morality  could  fail  to  seem  an  open  question.         SMITH'S  ANTI-­‐EXTERNALISM     Smith  rejects  the  possibility  of  principled  amoralism.    He  thinks  that  putative  amoralists  should   always  be  interpreted  as  making  moral  judgments  that  employ  moral  language  in  "inverted  commas"  –   they   are   expressing   indifference   not   to   the   moral   standards   they   accept   but   to   the   moral   standards   that   others   around   them   accept.     So   he   thinks   that   externalists   who   appeal   to   the   possibility   of   amoralism   are   begging   the   question.     Smith   defends   internalism   by   appeal   to   an   analogy   between   colors   and   morals:   motivation   is   essential   to   moral   judgment   in   the   way   that   qualitative   sensory   experience   is   essential   to   color   judgment.     Finally,   Smith   rejects   what   he   takes   to   be   the   externalist   explanation  of  the  way  in  which  the  motivation  of  the  "good  and  strong-­‐willed  person"  normally  tracks   changes  in  her  moral  beliefs.    He  believes  that  this  requires  a  de  dicto,  rather  than  a  de  re,  concern  to   do   the   morally   correct   thing.     This   commitment,   Smith   thinks,   is   implausible   and   commits   the   externalist  to  "one  thought  too  many"  (74-­‐75).     The   externalist   denies   that   an   agent's   motives   must   track   changes   in   her   moral   beliefs.     However,   even   the   externalist   should   claim   that   the   motives   of   the   "good   and   strong-­‐willed   person"   reliably  track  changes  in  her  moral  beliefs.    Smith  thinks  that  the  concerns  of  such  a  person  must  all  be   derived   from   her   moral   concerns.     This,   he   thinks,   is   inconsistent   with   intrinsic   concern   for   one's  

 

12  

associates;  the  externalist  must  ascribe  to  the  good  and  strong-­‐willed  person  "one  thought  too  many"   about  whom  to  save.    Externalism  leads  to  an  over-­‐moralized  self.     EXTERNALIST  REJOINDERS     There  are  several  places  the  externalist  may  want  to  put  her  foot  in  the  door.    Inverted  commas   readings   reject   a   literal   reading   of   the   putative   amoralist's   claims.     On   the   one   hand,   people's   self-­‐ reports  should  not  always  be  taken  at  face  value  (e.g.  Huck  Finn,  maybe).    But  why  should  we  always   reject  a  literal  reading?    What  if  the  putative  amoralist  persists  in  describing  his  judgments  as  moral   judgments  even  after  being  offered  an  inverted  commas  interpretation?     Why   does   Smith   think   that   the   appeal   to   principled   amoralists   question-­‐begging?     The   externalist   claim   is   not   that   the   internalist   does   not   have   a   consistent   position;   it   is   that   the   position   is   implausible.     Internalism   makes   a   very   strong   generalization   about   the   connection   between   moral   judgment   and   motivation,   and   externalism   challenges   that,   not   as   incoherent,   but   as   implausible.     Specifically,  it  is  implausible,  because  it  asserts  that  something  is  impossible,  which  seems  to  be  both   possible  and  actual.     In   appealing   to   principled   amoralism   as   a   reason   for   rejecting   (weak)   internalism,   the   externalist   is   reasoning   much   as   Smith   himself   does   in   appealing   to   unprincipled   amoralism   as   a   reason   for   rejecting   strong   internalism.     The   strong   internalist   could   preserve   consistency   by   insisting   that  the  apathetic  and  depressed  are  using  moral  language  in  inverted  commas.     Moreover,   in   cases   in   which   the   amoralism,   whether   principled   or   unprincipled,   is   acquired,   there   will   be   continuity   in   discrimination   and   belief   across   the   change   in   motivational   effect.     Since   we   interpreted  the  earlier  judgments  as  bona  fide  moral  judgments,  we  should  treat  the  later  judgments,   which  are  continuous  with  the  earlier  judgments,  as  bona  fide  moral  judgments.     The  charge  of  question-­‐begging  may  suggest  that  the  internalist  and  the  externalist  just  make   different  conceptual  claims  about  morality  and  the  question  then  becomes  which,  if  either,  conceptual   claim   to   accept,   and   this   may   seem   like   a   stalemate.     But   this   is   not   a   fair   description   of   the   debate.     Whereas   the   internalist   assumes   quite   a   bit   about   the   concept   of   morality   –   that   it   excludes   the   possibility  of  holding  beliefs  about  morality  and  beliefs  about  practical  reason  according  to  which  they   could   diverge   in   their   demands   –   the   externalist   assumes   much   less   about   the   concept   of   morality   –   in   particular,   the   externalist   can   recognize   a   wide   variety   of   possible   views   about   the   nature   and   content   of   morality   and   its   relation   to   practical   reason.     If   we   have   general   doubts   about   the   limits   of   conceptual   analysis   and   wish   to   be   able   to   entertain   very   diverse   conceptions   of   philosophical   concepts,  then  we  should  for  that  reason  prefer  externalism  to  internalism.     Now  consider  Smith’s  color  analogy.    We  might  wonder  if  the  color  analogy  begs  the  question   in  favor  of  internalism.    Why  should  we  find  an  analogy  between  moral  and  color  judgments  especially   apt  unless  we  already  accepted  internalism?    The  internalist  strategy  assumes  a  sort  of  dispositional   conception   of   colors,   which   is   itself   potentially   controversial.     But   even   if   we   make   this   assumption   about  colors,  there  are  disanalogies.    The  sort  of  sensory  experience  that  might  be  required  for  color   judgment  is  not  the  sort  of  conative  state  that  is  required  by  the  internalist  for  moral  judgment.    In  a   dispositional  analysis  of  color,  color  judgments  are  true  in  virtue  of  the  object  being  such  as  to  elicit  a   certain  sort  of  visual  experience  in  a  standard  observer.    But  it  won't  imply  that  anyone  making  a  color   judgment,   as   such,   must   have   this   visual   experience.     So,   it's   not   clear   that   the   moral   analog   of   this   could   deliver   the   desired   internalist   connection   between   moral   judgment   and   actual   motivation.     Also,   it  is  not  clear  that  the  capacity  for  visual  experience  is  a  condition  for  making  color  judgments.    Can't   the  blind  make  color  judgments,  perhaps  by  inheriting  meaning/reference  from  linguistic  participation   with  the  sighted  community?    Here,  we  may  need  to  distinguish  cases  of  congenital/acquired  blindness.     Now  consider  the  externalist  explanation  of  the  good  and  strong  willed  person.    Where  is  the   incompatibility   between   intrinsic   concern   for   associates   and   moralized   concern?     If   friendship   is   a   virtue,   then   won't   moralized   concerns   include   concerns   for   friends?     Perhaps   the   externalist   will   think   that   expressions   of   friendship   should   be   regulated   by   beliefs   about   the   permissibility   of   those   expressions.    Is  that  a  problem?    Don't  we  think  that  acting  on  concerns  for  friends  should  be  regulated  

 

13  

by  our  beliefs  about  what's  permissible?    It's  okay  to  favor  one's  spouse  in  a  rescue  situation  when  that   is  permissible,  but  not  when  it's  not  (e.g.  when  you  are  an  EMT).    It's  okay  to  favor  friends,  but  not  by   poisoning  their  rivals.     Smith’s  worry  about  the  externalist  seems  relevantly  like  Williamsesque  worries  about  Kant's   insistence   that   virtuous   agents   act   from   the   motive   of   duty.     Critics   often   complain   about   the   good   will   that   it   treats   sympathetic   or   compassionate   aid   as   morally   tainted.     But   this   misunderstands   the   doctrine  of  the  good  will.    The  good  will  does  not  require  contramoral  sentiments;  it  requires  only  that   they   be   regulated   by   a   sense   of   duty   such   that   the   agent   would   not   have   had   the   special   concern   or   would  not  have  acted  on  it  had  she  believed  the  action  in  question  was  impermissible.    Why  can’t  the   externalist  claim  about  the  motivation  of  the  good  and  strong-­‐willed  be  understood  in  a  similar  way?          

Moral Motivation and Moral Semantics - David O. Brink

Draft  of  10-­‐15-­‐14   PHIL  202:  Core  Ethics   Fall  2014;  Classics  in  Metaethics   David  O.  Brink   Handout  #2:  Moral  Motivation  and ...

189KB Sizes 0 Downloads 0 Views

Recommend Documents

No documents