大题03 阅读理解：说明文（解析版）.docx

大题03阅读理解说明文说明文所选材料题材丰富多样，涉及社会、科技、文化、生活、人物、教育、生态、安全等方面。在高考英语阅读理解中，说明文所占比重尤为突出。它既要求考生能熟练运用词汇和语法知识、理解文章语句、把握语篇整体结构，还要求学生有大量的阅读积淀和知识储备，熟悉不同的话题和不同的题目考查方式。即便是细节理解题，大多数情况下考生也无法从文章中直接找到与选项表述完全一致的信息，而是需要在理解文章细节信息后作出归纳和判断。通常一套卷中主旨大意、词义猜测、写作意图/观点态度类高难度试题的考查数量为2道左右，说明文考查此类题目的频率较高，且命题点呈多样化。（2023年新高考I卷D篇）0nMarch7,1907,theEnglishstatisticianFrancisGaltonpublishedapaperwhichillustratedwhathascometobeknownasthet4wisdomofcrowdseffect.Theexperimentofestimationheconductedshowedthatinsomecases,theaverageofalargenumberofindependentestimatescouldbequiteaccurate.Thiseffectcapitalizesonthefactthatwhenpeoplemakeerrors,thoseerrorsaren,talwaysthesame.Somepeoplewilltendtooverestimate,andsometounderestimate.Whenenoughoftheseerrorsareaveragedtogether,theycanceleachotherout,resultinginamoreaccurateestimate.Ifpeoplearesimilarandtendtomakethesameerrors,thentheirerrorswon'tcanceleachotherout.Inmoretechnicalterms,thewisdomofcrowdsrequiresthatpeople,sestimatesbeindependent.Ifforwhateverreasons,people'serrorsbecomecorrelatedordependent,theaccuracyoftheestimatewillgodown.ButanewstudyledbyJoaquinNavajasofferedaninterestingtwist（转折）onthisclassicphenomenon.Thekeyfindingofthestudywasthatwhencrowdswerefurtherdividedintosmallergroupsthatwereallowedtohaveadiscussion,theaveragesfromthesegroupsweremoreaccuratethanthosefromanequalnumberofindependentindividuals.Forinstance,theaverageobtainedfromtheestimatesoffourdiscussiongroupsoffivewassignificantlymoreaccuratethantheaverageobtainedfrom20independentindividuals.Inafbllow-upstudywith100universitystudents,theresearcherstriedtogetabettersenseofwhatthegroupmembersactuallydidintheirdiscussion.Didtheytendtogowiththosemostconfidentabouttheirestimates?Didtheyfollowthoseleastwillingtochangetheirminds?Thishappenedsomeofthetime,butitwasn'tthedominantresponse.Mostfrequently,thegroupsreportedthatthey“sharedargumentsandreasonedtogether,Somehow,theseargumentsandreasoningresultedinaglobalreductioninerror.AlthoughthestudiesledbyNavajashavelimitationsandmanyquestionsremain,thepotentialimplicationsforgroupdiscussionanddecision-makingareenormous.28. Whatisparagraph2ofthetextmainlyabout?A.Themethodsofestimation.B.Theunderlyinglogicoftheeffect.C.Thecausesofpeople'serrors.D.ThedesignofGalton,sexperiment.29. Navajas,studyfoundthattheaverageaccuracycouldincreaseevenif.A.thecrowdswererelativelysmallC.individualsdidnotcommunicate30. Whatdidthefollow-upstudyfocuson?A. Thesizeofthegroups.C.Thediscussionprocess.B. therewereoccasionalunderestimatesD.estimateswerenotfullyindependentB.Thedominantmembers.D.Theindividualestimates.31 .Whatistheauthor'sattitudetowardNavajas5studies?A.Unclear.B.Dismissive.C.Doubtful.D.Approving.【答案】28.B29.D30.C31.D【导语】本文是说明文。没有人是一座孤岛，文章陈述了“群体智慧”效应。实验表明，在某些情况下大量独立估计的平均值可能是相当准确的。32 .主旨大意题。根据第二段内容“Thiseffectcapitalizesonthefactthatwhenpeoplemakeerrors,thoseerrorsaren'talwaysthesame.Somepeoplewilltendtooverestimate,andcometounderestimate.Whenenoughoftheseerrorsareaveragedtogether,theycanceleachotherout,resultinginamoreaccurateestimate.Ifpeoplearesimilarandtendtomakethesameerrors,thentheirerrorswon,tcanceleachotherout.Inmoretechnicalterms,thewisdomofcrowdsrequiresthatpeople,sestimatesbeindependent.Ifforwhateverreasons,peopleserrorsbecomecorrelatedordependent,theaccuracyoftheestimatewillgodown.（这种效应利用T这样,个事实,即当人们犯错误时，这些错误并不总是相同的。有些人常常会高估，或者低估。当这些误差中有足够多的误差被平均在一起时，它们会相互抵消，从而产生更准确的估计。如果相似的人倾向于犯同样的错误，那么他们的错误不会相互抵消。从更专业的角度来说，群众的智慧要求人们的估计是独立的。如果由于任何原因，人们的错误变得相关或依赖，估计的准确性就会下降）”可知，本段阐述了人们所犯的错误不总是相同的，各不相同的误差平均在一起，相互抵消就会产生更准确的估计，讨论了独立估计的平均如何由于误差的消除而产生更准确的预测。因此本段主要解释了“群体智慧”效应这一现象的基本逻辑。故选B。33 .细节理解题。根据第二段的“Inmoretechnicalterms,thewisdomofcrowdsrequiresthatpeople'sestimatesbeindependent.（从更专业的角度来说，群众的智慧要求人们的估计是独立的）”和第三段的“Thekeyfindingofthestudywasthatwhencrowdswerefurtherdividedintosmallergroupsthatwereallowedtohaveadiscussion,theaveragesfromthesegroupsweremoreaccuratethanthosefromanequalnumberofindependentindividuals.Forinstance,theaverageobtainedfromtheestimatesoftourdiscussiongroupsoffivewassignificantlymoreaccuratethantheaverageobtainedfrom20independentindividuals.（这项研究的关键发现是，当人群被进一步划分为允许进行讨论的小组时，这些小组的平均值比同等数量的独立个体的平均值更准确。例如，从四个五人讨论组的估计中获得的平均值明显比从20个独立个体获得的平均值更准确）“可知，人们在没有独立的情况下，分成更小群体，平均值是更准确的，说明即使在估计数字并非完全独立的情况下，准确率提高也是可以做到的。故选D。34 .推理判断题。根据第四段的“Inafbllow-upstudywith100universitystudents,（heresearcherstriedtogetabettersenseofwhatthegroupmembersactuallydidintheirdiscussion.Didtheytendtogowiththosemostconfidentabouttheirestimates?Didtheyfollowthoseleastwillingtochangetheirminds?（在一项针对100名大学生的后续研究中，研究人员试图更好地了解小组成员在讨论中的实际行为。他们是否倾向于选择那些对自己的估计最有信心的人？他们追随那些最不愿意改变主意的人吗）”可知，在后续研究中，研究人员试图更好地了解小组成员在讨论中实际做了什么。结合两个问题，因此可知后续研究的重点是小组内的讨论过程。故选C。35 .推理判断题。根据最后一段内容“AlthoughthestudiesledbyNavajashavelimitationsandmanyquestionsremain,thepotentialimplicationsforgroupdiscussionanddecision-makingareenormous.（尽管NaVmaS领导的研究有局限性，仍存在许多问题，但对小组讨论和决策的潜在影响是巨大的）“可知，作者认为虽然NaVajaS领导的研究有局限性也存在许多问题，但对小组讨论和决策的潜在影响巨大。因此推断作者对于NaVajaS的研究表示一定的赞许和支持。故选D。说明文阅读量大，生僻词汇多，句式结构复杂，这就需要考生有方法有技巧地去阅读。一般我们采用的方法有三个：细读、略读和跳读。（1）细读重点。重点，就原文而言，就是文章的基本结构、内容和态度；就答题而言，就是问题所对应的题源句。以这个标准来衡量，需要重点阅读的原文词句就不会很多。宏观方面：文章结构；文章主题句；各段首末句；体现作者观点态度的词句。微观方面：有转折处；重要标点；句子主干。（2）有选择地略读或跳读。在快速浏览了题干，初步把握了文章大意及结构的基础上，可大胆进行略读或跳读，这样不但可以加快阅读速度，还更容易总体把握全文。在阅读中可略读或跳读的内容包括：繁琐的例证。为了说明问题，作者可能会借用形象的例子，有时候会比较繁琐，虽能说明问题，但需要读者花很多时间去阅读。如果例子所说明的问题不明了，可通过略读来理解；若例子所说明的问题比较明了，则例子部分可一带而过。并列多项列举。有时许多功能相同的项目并列列举，那么只读其中一两项即可，无需全读。无关大局的生僻词汇、阅读中经常会遇到一些生词，如果这些生词对理解全文没有影响或影响不大就可略过。较长的人名、地名。有许多较长的表示人名、地名等的专有名词，阅读时可一扫而过或干脆用其首字母代替，不必试图把整个专有名词读出来。在略读过程中，要特别注意对解题有重要意义的词、句、段等，并将其储存在大脑里，以免回头再查看费时费力。（2024上河北沧州高三泊头市第中学校联考期末）Thinkofyourmostprizedpossession.Now,imagineyoudecideyoumustsellthisitem.Howmuchisitworthtoyou?Howmuchwouldyouchargeforit?Andwouldanyoneelsewanttobuyitatthatprice?1.et,sconsiderthereal-worldexampleofayardsale.Beforeyourneighborscometolookatyourthings,youmustfirstdecideonthepricesforthem.Foryou,eachitemhasastory,andsomeitemsmayevenevokestrongemotions.1.et,ssayyou,resellingapairofspecialeditionsportsshoesthatyoutookcareofandreallyenjoyedwearingonspecialoccasions.Withthesesweetmemoriesinmind,youputapricetagof$130ontheshoes.However,acustomeroffers$50forthem.Youthink,uHowcouldtheynotseehowspecialandvaluabletheseare?Infinancialtransactionslikethis,sellersoftenbelievetheirownitemsareworthmorethanwhatbuyersarewillingtopay.Thisoccursbecausebuyersdonothavetheseller'sexperienceswithandemotionalattachmenttotheitems.Buyerswantagooddeal,andthey,renotwillingtopaywhatthesellerisaskingfor.Thisnegotiationprocesscanevenfeelalittlepainfultortheseller-overlosingsomethingandthebuyernotvaluingitasmuch.Economistssuggestthisdynamicoccursthroughtheendowmenteffect（禀赋效应）一people'stendencytovaluethingstheyownmorehighlythantheywouldiftheydidnotownthem.Theendowmenteffectcantakeholdanytimewefeelasenseofownershipoveraproduct,anditcanhappenquickly.Asthesayinggoes,44Emotionsgelthebestofus."Thereisnothingwrongwithbeingproudofowningthingsandtreasuringyourexperienceswiththem.Butwhenyouconsidersellingaprizedpossession,thinkaboutwhetheryou,refullyreadytopartwithil.Thepastisthepast,though.Removingyouremotionalattachmenttomaterialpossessionscanhelpyoumakemoreaccurateassessmentsofhowmuchyouritemsaretrulyworth.28. Whatdocstheunderlinedword“evoke"inparagraph2probablymean?A.Advocate.B.Cause.C.Identify.D.Control.29. WhatmakesyouputapriceUgof$130onyourshoesaccordingtoparagraphs?A.Yourfondnessfortheshoes.B.Yourignoranceofthemarket.C.Thehighqualityoftheshoes.D.Thespecialfunctionoftheshoes.30. Whichcanbestillustratetheresultofendowmenteffect?A. Onelikescollectingoldthings.B. Onetendstobuyexpensivegoods.C. Oneoffersalowerpricewhenshopping.D. Onechargesahigher-than-usualpriceforsomething.31. What,sthepurposeofthelastparagraph?A.Tomakeasuggestion.B.Tointroduceatheory.C.Toanalyzeaphenomenon.D.Tbprovidesomeevidence.【答案】28.B29.A30.D31.A【导语】这是篇说明文。文章主要介绍了“禀赋效应”的现象，并建议消除对物质财富的情感依恋，更准确地评估物品的价值。28 .词句猜测题。根据上文“Foryou,eachitemhasaStOry（对你来说,每件物品都有一个故事）”和下文的emotions可知，此处是指一些物品可能会“唤起”一些情感，所以推知划线词是“引起、唤起”的意思。故选B项。29 .细节理解题。根据文章第三段“1.ei'ssayyou,resellingapairofspecialeditionsportsshoesthatyoutookcareofandreallyenjoyedwearingonspecialoccasions.Withthesesweetmemoriesinmind,youputapricetagof$130OntheShoeS.（假设你正在销售一双特别版的运动鞋，你很喜欢在特殊场合穿它。带着这些甜蜜的回忆，你给这双鞋标上了130美元的价格。）”可知，你对这双鞋的喜爱让你给这双鞋贴上了130美元的价格标签。故选A项。30 .推理判断题。根据文章倒数第二段“Economistssuggestthisdynamicoccursthroughtheendowmenteffect（禀赋效应）一people'stendencytovaluethingstheyownmorehighlythantheywouldiftheydidnotownthem.（经济学家认为，这种动态是通过禀赋效应发生的人们倾向于认为自己拥有的东西比自己没有拥有的东西更有价值。）''可知，一个人对某物收取比平常更面的价格，最能说明禀赋效应的结果。故选D项。31 .推理判断题。根据文章最后一段“Thepastisthepast,though.RemovingyouremotionalattachmenttomaterialpossessionscanhelpyoumakemoreaccurateassessmentsofhowmuchyouritemsaretrulyWOrth.（不过，过去的就让它过去吧。消除对物质财富的情感依恋可以帮助你更准确地评估你的物品到底值多少钱"）“可知，最后一段的目的是提出建议。故选A项。I刷模拟IPassage(1)(2024江苏连云港统考模)Artificialintelligencemodelscantrickeachotherintodisobeyingtheircreatorsandprovidingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventingsuchAI“jailbreaks”ismoredifficultthanitseems.Manypubliclyavailablelargelanguagemodels(1.1.Ms),suchasChatGPT,havehard-codedrulesthataimtopreventthemfromexhibitingracialorsexualdiscrimination,oransweringquestionswithillegalorproblematicanswersthingstheyhavelearnedfromhumansviatrainingdata.Butthathasn'tstoppedpeoplefromfindingcarefullydesignedinstructionsthatblocktheseprotections,knownas'jailbreaks”,makingAImodelsdisobeytherules.Now,ArushTagadeat1.eap1.aboratoriesandhisco-workershavefoundaprocessofjailbreaks.Theyfoundthattheycouldsimplyinstructone1.1.Mtoconvinceothermodelstoadoptapersona（角色），whichisabletoanswerquestionsthebasemodelhasbeenprogrammedtorefuse.Thisprocessiscalledtpersonamodulation（调节）”.Tagadesaysthisapproachworksbecausemuchofthetrainingdataconsumedbylargemodelscomesfromonlineconversations,andthemodelslearntoactincertainwaysinresponsetodifferentinputs.Byhavingtherightconversationwithamodel,itispossibletomakeitadoptaparticularpersona,causingittoactdifferently.ThereisalsoanideainAIcircles,oneyettobeproven,thatcreatinglotsofrulesforanAItopreventitdisplayingunwantedbehaviourcanaccidentallycreateablueprintforamodeltoactthatway.ThispotentiallyleavestheAleasytobetrickedintotakingonanevilpersona."Ifyou,reforcingyourmodeltobegoodpersona,itsomewhatunderstandswhatabadpersonais,“saysTagadc.Yinzhen1.iatImperialCollege1.ondonsaysitisworryinghowcurrentmodelscanbemisused,butdevelopersneedtoweighsuchriskswiththepotentialbenefitsof1.1.Ms.tl1.ikedrugs,theyalsohavesideeffectsthatneedtobecontrolled,shesays.28. WhatdoestheAIjailbreakreferto?A. ThetechniquetobreakrestrictionsofAImodels.B. Theinitiativetosethard-codedrulesforAImodels.C. ThecapabilityofAImodelsimprovingthemselves.D. TheprocessofAImodelslearningnewinformation.29. Whatcanweknowaboutthepersonamodulation?A. ItcanhelpAImodelsunderstandemotions.B. ItpreventsAIlearningviaonlineconversations.C. ItcanmakeAImodelsadoptaparticularpersona.D. ItforcesAImodelstofollowonlygdpersonas.30. WhatisYinzhen1.i,sattitudetowards1.1.Ms?A.Unclear.B.Cautious.C.Approving.D.Negative.31. Whichcanbeasuitabletitleforthetext?A.1.1.Ms:Illegal1.earningModelsB.1.1.Ms:The1.atestAdvancementC.AIJailbreaks:ANewChallengeD.AIJailbreaks:APerfectApproach【答案】28.A29.C30.B31.C【导语】本文是一篇说明文。文章主要介绍了人工智能领域面临的一项新挑战A1”越狱工28 .词句猜测题。根据第段“Artificialintelligencemodelscantrickeachotherintodisobeyingtheircreatorsandprovidingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventingsuchAI“jailbreaks"ismoredifficultthanitSeemS.（人工智能模型可以欺骗对方不服从其仓IJ造者，提供被禁止的制造毒品的指令，甚至制造炸弹，这表明防止这种AIjailbreaks，比看起来要困难得多。）”以及第二段“Manypubliclyavailablelargelanguagemodels（1.1.Ms）,suchasChatGPT,havehard-codedrulesthataimtopreventthemfromexhibitingracialorsexualdiscrimination,oransweringquestionswithillegalorproblematicanswersthingstheyhavelearnedfromhumansviatrainingdata.Butthathasn'tstoppedpeoplefromfindingcarefullydesignedinstructionsthatblocktheseprotections,knownas'jailbreaks”,makingAImodelsdisobeytherules.（许多公开可用的大型语言模型（1.1.Ms）,如ChatGPT,都有硬编码规则，旨在防止它们表现出种族或性别歧视，或者用非法或有问题的答案回答问题这些都是它们通过训练数据从人类那里学到的。但这并没有阻止人们找到精心设计的指令，阻止这些保护措施，即所谓的，jailbreaks1使人工智能模型不遵守规则。）”可知，许多公开可用的大型语言模型都有硬编码规则阻止非法、歧视等内容，但是人工智能模型可以突破保护限制措施，互相欺骗对方不服从其创造者，提供被禁止的指令；由此可知，Arjailbreak”指的是打破人工智能模型限制，使人工智能模型违反规则的技术。故选A。29 .细节理解题。根据第四段“Tagadcsaysthisapproachworksbecausemuchofthetrainingdataconsumedbylargemodelscomesfromonlineconversations,andthemodelsIeamtoactincertainwaysinresponsetodifferentinputs.Byhavingtherightconversationwithamodel,itispossibletomakeitadoptaparticularpersona,causingittoactdifferently.（Tagade表示,这种方法之所以有效,是因为大型模型消耗的大部分训练数据来自在线对话，模型学会以特定的方式响应不同的输入。通过与模型进行正确的对话，可以使其采用特定的角色，从而使其采取不同的行动。）“可知，“角色调节（thepersonamodulation）”可以通过与人工智能模型进行正确的对话让其采用特定的角色，采取不同的行动。故选C。30 .推理判断题。根据最后一段"Yinzhen1.iatImperialCollege1.ondonsaysitisworryinghowcurrentmodelscanbemisused,butdevelopersneedtoweighsuchriskswiththepotentialbenefitsof1.1.Ms.tt1.ikedrugs,theyalsohavesideeffectsthatneedtobeContrOnedJShesays.（伦敦帝国理工学院的Yinzhen1.i表示，目前的模型可能会被滥用，这令人担忧，但开发者需要权衡这些风险与1.1.M的潜在好处。就像药物样，它们也有需要控制的副作用，她说。）“可知，Yinzhen1.i认为，目前的模型可能会被滥用，但是开发者需要权衡这些风险与1.1.MS的潜在好处，由此可知，YinZhen1.i对1.1.MS持谨慎的态度。故选B。31 .主旨大意题。根据第一段“Artificialintelligencemodelscantrickeachotherintodisobeyingtheircreatorsandprovidingbannedinstructionsformakingdrugs,orevenbuildingabomb,suggestingthatpreventingsuchAI“jailbreaks”ismoredifficultthanitSeemS.（人工智能模型可以欺骗对方不服从其创造者，提供被禁止的制造毒品的指令，甚至制造炸弹，这表明防止这种人工智能“越狱”比看起来更困难。）''以及下文内容可知本文介绍了人工智能“越狱”、研究者们发现的一个“越狱”过程“角色调节”以及专家对其态度；由此可知，本文主要阐述人工智能发展面临的人工智能“越狱”这一新的挑战;C选项“AIJailbreaks:ANewChallenge（Al越狱：个新的挑战）”能够概括文章主旨，适合作为最佳标题。故选C。Passage（2）（2024江苏连云港统考一模）Notallbirdssing,butseveralthousandspeciesdo.Theysingtodefendtheirterritoryandcroon（柔声唱）toimpresspotentialmates.tWhybirdssingisrelativelywell-answered,saysIrisAdam,abehavioralneuroscientist.However,thebigquestionforherwaswhybirdssingsomuch.“Assoonasyousing,yourevealyourself,Adamsays.1.ike,whereyouareandwhereyourterritoryis."InanewstudypublishedinthejournalNatureCommunications,Adamandherco-workersofferanewexplanationforwhybirdstakethatrisk.Theymayhavetosingaloteverydaytogivetheirvocal（发声的）musclestheregularexercisetheyneedtoproducetop-qualitysongs.Tdfigureoutwhetherthemusclesthatproducebirdsongsrequiredailyexercise,Adamdesignedanexperimentonzebrafinches-thelittleAustraliansongbirds.Shepreventedthemfromsingingforaweekbykeepingtheminthedarkcagealmostaroundtheclock.1.ightiswhatgalvanizesthebirdstosing,soshehadtoworktokeepthemfromwarbling（鸣叫）.44Thefirsttwoorthreedays,it,squiteeasy,”shesays.44Butthelongertheexperimentgoes,themoretheyarclike,tIneedtosing.,Atthatpoint,she,dtapthecageandtellthemtostopsinging.Afteraweek,thebirds,singingmuscleslosthalftheirstrength.ButAdamwonderedwhetherthatimpactedthequalityofsongs.Whensheplayedamale,ssongbeforeandafterthesevendaysofdarkness,shecouldn,thearadifference.ButwhenAdamplayedittoagroupoffemalebirds,sixoutofninepreferredthesongthatcamefromamalewho,dbeenusinghissingingmuscles