<s>
The_AT scientist_NN1 has_VHZ the_AT opportunity_NN1 to_TO advance_VVI human_JJ knowledge_NN1 in_II a_AT1 very_RG special_JJ way_NN1 ._. 
</s>
<s>
His_APPGE or_CC her_APPGE discoveries_NN2 may_VM open_VVI new_JJ vistas_NN2 for_IF humans_NN2 ,_, for_IF the_AT development_NN1 of_IO society_NN1 ,_, and_CC for_IF the_AT protection_NN1 of_IO nature_NN1 ._. 
</s>
<s>
Whether_CSW one_PN1 studies_VVZ biology_NN1 ,_, biochemistry_NN1 ,_, chemistry_NN1 ,_, mathematics_NN1 ,_, physics_NN1 ,_, or_CC medicine_NN1 ;_; or_CC any_DD other_JJ science_NN1 ,_, his_APPGE or_CC her_APPGE role_NN1 is_VBZ always_RR a_AT1 very_RG important_JJ one_PN1 ._. 
</s>
<s>
Sometimes_RT ,_, this_DD1 role_NN1 is_VBZ so_RG important_JJ that_CST it_PPH1 provides_VVZ life_NN1 and_CC death_NN1 answers_NN2 ._. 
</s>
<s>
Because_CS I_PPIS1 do_VD0 n't_XX know_VVI enough_RR about_II the_AT sciences_NN2 ,_, I_PPIS1 want_VV0 to_TO know_VVI that_CST I_PPIS1 can_VM trust_VVI scientists_NN2 so_CS21 that_CS22 I_PPIS1 do_VD0 n't_XX have_VHI to_TO worry_VVI that_CST they_PPHS2 may_VM seek_VVI to_TO deceive_VVI me_PPIO1 about_II such_DA matters_NN2 ._. 
</s>
<s>
At_II the_AT same_DA time_NNT1 ,_, we_PPIS2 should_VM always_RR be_VBI aware_JJ that_CST some_DD scientists_NN2 do_VD0 provide_VVI false_JJ evidence_NN1 to_TO support_VVI faulty_JJ claims_NN2 ._. 
</s>
<s>
This_DD1 can_VM lead_VVI to_II great_JJ danger_NN1 for_IF us_PPIO2 ._. 
</s>
<s>
The_AT key_NN1 to_II scientific_JJ success_NN1 lies_VVZ in_II the_AT scientific_JJ attitude_NN1 ,_, which_DDQ means_VVZ that_CST one_PN1 tests_VVZ over_RP and_CC over_II hypotheses_NN2 ,_, and_CC conducts_VVZ many_DA2 experiences_NN2 to_TO be_VBI sure_JJ that_CST the_AT results_NN2 of_IO the_AT study_NN1 prove_VV0 the_AT hypotheses_NN2 with_IW validity_NN1 and_CC reliability_NN1 ._. 
</s>
<s>
That_REX21 is_REX22 ,_, scientists_NN2 ask_VV0 questions_NN2 which_DDQ are_VBR educated_JJ guesses_NN2 ,_, backed_VVD up_RP by_II an_AT1 understanding_NN1 of_IO former_DA or_CC similar_JJ studies_NN2 ,_, and_CC the_AT scientists_NN2 want_VV0 to_TO know_VVI whether_CSW their_APPGE results_NN2 can_VM be_VBI considered_VVN true_JJ ,_, based_VVN on_II their_APPGE evidence_NN1 and_CC data_NN which_DDQ they_PPHS2 have_VH0 collected_VVN ._. 
</s>
<s>
If_CS the_AT results_NN2 are_VBR true_JJ once_RR ,_, they_PPHS2 also_RR want_VV0 to_TO know_VVI if_CSW the_AT hypotheses_NN2 can_VM be_VBI tested_VVN again_RT and_CC produce_VV0 similar_JJ ,_, thus_RR reliable_JJ ,_, results_NN2 ._. 
</s>
<s>
If_CS scientists_NN2 make_VV0 faulty_JJ assumptions_NN2 ,_, or_CC use_VV0 incorrect_JJ evidence_NN1 ,_, or_CC produce_VV0 untrue_JJ results_NN2 from_II their_APPGE studies_NN2 ,_, then_RT not_XX only_RR will_VM their_APPGE reputation_NN1 be_VBI damaged_VVN ,_, but_CCB also_RR the_AT program_NN1 in_II which_DDQ they_PPHS2 work_VV0 ,_, and_CC possibly_RR a_AT1 whole_JJ field_NN1 of_IO science_NN1 will_VM suffer_VVI ._. 
</s>
<s>
Honesty_NN1 is_VBZ a_AT1 matter_NN1 of_IO self_NN1 interest_NN1 for_IF scientists_NN2 ,_, because_CS a_AT1 scientist_NN1 's_GE reputation_NN1 for_IF integrity_NN1 is_VBZ his_APPGE or_CC her_APPGE most_RGT valuable_JJ asset_NN1 ._. 
</s>
