What is a Good Argument? Validity and Truth - Marianne Talbot

okay here we are at week four well done for staying with us so far this week we're going to be looking at how to evaluate arguments how to tell whether an argument is a good one or a bad one and we'll start with inductive arguments right let's get started today last week we learned how to analyze arguments and what I meant by that was how to identify them and how to set them out logic book style I gave you six steps to analyzing an argument and these are the only steps another thing that's become clear to me from emails and questions I've had this week is that a lot of people are trying to evaluate an argument to say whether it's a good argument or a bad argument as they try and analyze it well don't because you'll always be led astray if you try to do that and especially with the complicated arguments like the one we looked at last week so follow just these steps don't do anything else to the argument don't say I think the conclusion shouldn't have a knot in it here and take the knot out knots are really very important and they shouldn't be added in either if they're not there and so so just one of these steps that's all you need to do and I'm not suggesting it's easy in fact this is the hardest thing you'll ever do in logic Computers can't do this only we can do this a computer can evaluate arguments very easily by appeal to just a very simple algorithm but what it can't do is translate an argument in English into a formal language hopeless computers can't do that or at least not unless they're very very very simple so those are the those are the steps that you must take to analyze arguments and don't try and evaluate them at the same time and okay we did see that although we needed to paraphrase arguments in order to complete these steps in other words we had to add in things that I mean instead of it we put she or something like that all that wasn't a good example but instead of it it was tried to tickle him or do you remember so we had to paraphrase arguments to complete this assessment by paraphrase I just mean put what's there in different words not change the meaning of anything and certainly don't add in any meanings or take away any meanings paraphrase is just changing them the words so that the argument structure becomes clearer do you see the difference and again you'll probably need a bit of practice before that comes easily because it really is a temptation to evaluate the arguments and to change its meaning if you think it would be clearer if so-and-so said this rather than that but try to avoid that because what you're trying to do is identify the argument somebody else is making not the argument that you would make if you were in his position okay this is the point about analyzing arguments is in the hope that you might learn something and you won't do that if you're imposing your own grid of understanding onto someone else's argument okay so paraphrase but don't change the meaning we also saw that it's necessary to bring to bear our understanding of the arguments for example do you remember the suppressed premises that we added last week I mean we had quite a tussle with some of them didn't we some of them turned out some of the things that we thought might be suppressed premises turned out actually to be a matter of inconsistent terms or something like that so we have to bring to bear our understanding of the arguments and what follows from that but don't read into the argument anything that isn't actually there if a suppressed premise is there it's really pretty clear that that's a suppressed premise of the argument it's a premise that ought to be there but isn't so all you're doing is making it explicit something that's already there implicitly okay I think we're okay and and I've just said it's extremely important to in analyzing an argument not to evaluate it first you identify it then you evaluate it okay any questions about all that before I move on to today nope okay let's move on to today what we're going to do today is to start learning how to evaluate arguments now today I've got down there starting with validity and truth looking at the distinction between them but I've decided instead to start with induction and then go on to validity in truth next week and then look at deductive arguments and the evaluation of them in the final week so we're going to deal with induction this week oh okay oh I've done it now it's going to ask you to tell me what an inductive argument was but there we are okay you knew this anyway didn't you yes good okay they fantastic inductive arguments is such that the truth of their premises makes the truth of the conclusion more or less likely okay and if you remember we looked at two examples in the first place we looked at the sun's rising the Sun has risen every day in the history of the world therefore the Sun will rise tomorrow and every time you see Mary Ann she's been wearing earrings so next time you see her she'll be wearing earrings I'm gonna leave them off next week if I remember all inductive arguments rely on the principle of the uniformity of nature as human called it David Hume called it and and the only arguments for the principle of the uniformity of nature itself are themselves inductive so it looks as if any argument you offer for induction is going to be circular and based on induction itself and this is a this is a real problem people would love to be able to justify the principle of the uniformity of nature to say why we should believe that the future will be like the past but no one's conclusively succeeded there's reams and reams and reams books and papers written on on this problem and there are lots of theories about it but there's no theory on which everyone would converge yet okay different types of inductive argument inductive generalizations calls all generalizations arguments from analogy and/or arguments from Authority we're going to have a look at each of these separately and look at how to evaluate them so by how to evaluate how to tell whether they're good arguments or bad arguments because remember inductive arguments are not it's not a matter of either all with inductive arguments they're either strong or weak okay so there's a gradation it's a matter of degree as to how good an inductive argument is okay let's start with inductive generalizations and what I mean by this is that the premise identifies the characteristic of a sample of the population of a population and the conclusion extrapolates that characteristic to the rest of the population and all inductive arguments are actually a form of this of inductive generalization so in learning how to evaluate inductive generalizations you can apply everything you learn to other types of inductive generalization but let's have a look at them generally okay here are two examples so okay looking first at this one what's the population that we're looking at here so do you remember I said the premise points to a sample of a population and the conclusion extrapolates to the rest of the population so what do I mean by the population in this case voters exactly that's right so we're saying here that 60% of the voters have been sampled and that 60% said they'd vote for mr. many promise and we're extrapolating from that to therefore actually there's a suppressed premise here isn't there or there's something we could add in here well no we'll move on to that in a minute we we we're sort of assuming aren't we that 60% of the of the population as a whole would be enough for him to win do you see what I mean because that's implied by this isn't it rather than actually stated okay and then the other one we've got here what's the sample sorry what's the population here one one calls to BT yep calls to BT so the premise says whenever I've tried to ring BT whenever I've tried to make calls to BT it's taken me hours and I'm extrapolating to that for that too it's called by me actually isn't it rather than calls by calls generally so I'm extrapolating from my past experience to my future experience correctly and okay so what I want you is to have a look at each of these arguments or you can choose just one of them if you want to do it more slowly and us you're certain to write down the questions to which you would need answers in order to decide whether these are good arguments and then we'll go through them together so have a look yourself and just think about these questions and think about what you would ask in order to satisfy yourself that these were good arguments okay anyone want to give me examples those sort of questions you would ask okay if if the electorate was just ten people why would that help you evaluate the argument it is it really the population the number of the population you want or what else might it be you might want to know the size of the sample yeah yes I thought you might because if you've got ten people only we're in the sample and yet there are million people in the population then the sample just isn't big enough is it I've gotta know it isn't implicit in the word voters you might want yes I mean one of the things you would certainly want to consider here is that the voters sampled said that they would vote for mister what is it whatever his name is but actually won't vote for him or may not vote at all yes I mean either way it wouldn't make much difference so yes I don't think that's a yes that's a bit of background information that you would bring to bear on this particular argument something you know about voters which show that you you really have to know a bit more about well you present presumably I bet it expects there's a number by which they determine how many are likely to actually say I don't know yes you would say you'd certainly need to know whether they were telling the truth yep okay okay it's certainly the case that mr. many promise is not likely to win if he's not going to stand even if 60% of the votes so actually that's quite a good cancer example isn't it a case where the prayer the premise would be true but the conclusion would have to be false that is quite a good counter example to that if you got situation where the voters really did want to vote for whoever it was but he wasn't you understand yeah I like that one another one here good you'd want to know whether the sample is representative wouldn't you because if the only people they asked were males then who knows what women are going to do or if they're all under 24 or if they're all black or if they're all whatever you need to know that the sample chosen is representative of the population as a whole yes okay so if you have something like the radio there was a radio program wasn't there that was taking votes for something rather and a lot of people so I mean actually what you want to do is you want to ask whether the premise here is true at all yes definitely yep yes because if six maybe that 60% of the voters said that they vote for mr. brown but then something dreadful happens and and it's certainly not the case that if you sampled them again just before the election they would still say good you're coming up with all sorts of things I haven't got myself here this is brilliant who did the sampling yep that that would be a very good thing and again I mean that's another example of is the premise true because if the person saying that 60% of the voters said that they would vote for him if they're all apparatchiks for mr. many promise who want to make him feel good before the election you might question the premise itself might and you okay what about this one or is there anything that would be added to this one that we haven't already considered gentlemen there when I think is perfectly good because if I've been trying to ring Beatty at two o'clock in the morning it might be perfect you know yes it may have taken hours but where I - ring at 10 o'clock in the morning it might be different I'm assuming that they don't answer the phone at two o'clock in the morning okay it's certainly reasonable to ask whether it it's just me yes I mean there might be something about my particular telephone number that whenever I bring bt there's something that says don't answer this one or something like that but as the conclusion is that when I ring bt do you see what I mean again this this again the way I've set this up the population here is calls that I make to bt rather than calls that anyone makes to bt yes I might have only made one or two I mean again that's structurally the same as when we said here how many people did we sample in the population and what percentage of that elder the population is that and you're suggesting exactly the same thing here quite properly if I've only tried to ring once or twice then you know is that really a big enough sample good again you're questioning whether that's premise is true I mean maybe I'm just very bad at calculating time maybe I'm one of these people who's very keen to get somebody answer my phone call immediately and if it takes 30 seconds then I get very irritated and thinks it's I think it's ours okay you would have to assume wouldn't it that that it was the same part of BT again because otherwise otherwise you get an equivocation wouldn't you as bt here wouldn't mean the same as bt here okay an equivocation by the way is there is an argument in which you use the same word with two different meanings okay so if you think of the word Bank it could mean financial institution it could mean an action of an aeroplane or it can mean the side of a river and if in an argument you used it in all three of those meanings you can imagine an argument that would look good but as a matter of fact wouldn't work at all and that's as a result of equivocation you're equivocating on the word bank so if I were equivocating here on the word bt or the the letters bt my conclusion might not follow from my premises okay very good that that really is good I think it's very impressive you'll see as I go through the things that I'm going to list that you you've said just about all of them okay firstly is that just about all of them there's one I think I've got that you haven't it's the premise true okay we've got sixty percent of the samples said that they would vote for mister many promise well can we really believe that might they be bad at record-keeping so it actually wasn't sixty percent it was only 50 percents and you know if you you last year when you use those people they were completely hopeless might they be engaged in wishful thinking might they be bad just bad at maths they can't work out percentages and am I telling the truth am I in the paid pay of one of BT's rivals am i prone to exaggeration am I just very bad at estimating time so lots of reasons why the premise itself might not be true and if you remember whenever we're evaluating an argument there are two things we've got to look at can you remember what they are just two basic things we look at whenever we are evaluating an argument of any kind at all one is does the conclusion follow from the premises that's right and the other is are the premises true that's right is if even one premise is false then then that doesn't guarantee the truth of the conclusion does it or doesn't even make the truth of the conclusion more likely so first thing you look at when you look at any arguments is are the premises true okay how large is the sample again you got this how many of those who would vote in the election were sampled 10 out of 1 million well that doesn't look very good does it a thousand out of 1 million that looks better how many is enough though do you think and that's a really difficult question isn't it how many is enough I I'm just specifying here that 1 million is the population and then we're saying ok how many of those would count as enough and I'm saying there actually isn't any answer to that we can certainly answer that 10 is probably not enough and we might be able to say that 999 thousands or ten that I don't even know how much a million is a thousand thousand isn't it okay nine hundred thousand would be enough okay but in between those two numbers what counts as enough well that's coming later that's coming when we look at the representativeness of the sample at the moment the only thing we're talking about is the size of the sample if I say all swans are white and you say well what's your reason for saying that and I say well I saw a swan just now and it was white and you say what just one and I see yeah and you can be more or less inductively bold and actually if we were to look at people in this room if we were to do a head count of people in this room we'd find that some of us are very oh I shouldn't say us because I'm not inductively bold but some of us would be prepared to extrapolate from a very small number and others of us would be very skeptical about extrapolating even from quite a large number so actually the question how many is enough the answer would be it depends on who you are on and on how inductively bold you are not well statistics additions have to come up with a something that they would count as enough when you say the larger the sample do you mean that it's certainly true that if the thousands have been samples that's much more confidence boosting than 10 that is that's what you mean yes yes okay no I think you know I'm getting out of my depth here I don't I don't understand what you're saying I'm afraid yes we're coming to that that's representativeness no let's let's leave representativeness aside at the moment I'm just taught at the moment I'm just talking about size and all we need to look at is how many of those in the sample how many in the population how many in the sample do we think that we've got enough who've been sampled in order to make us more confident about the the extrapolation if I've only rung BBT once then my claim that the next time I'm going to ring is is really pretty low isn't it it's a very weak argument whereas if I've rung BT 50 times and not got through then that's more reason to think so if we think remember that inductive arguments make the premises make the conclusion more or less likely well if my premises I've rung BT once in the past and it took them hours to answer then so it'll take them hours to answer again my arguments much less strong than if I say I've rung bt-50 odd times in the past and they've taken them hours to answer therefore it'll take me hours to answer next time see what I mean and again here if I say 10 out of 160 percent of 10 in other words 6 voters out of a million said that they'd be voting for mister many promise therefore mister many promise will win the election that's a less good argument a weaker argument than if I say 60% of a thousand voters say that they'll votes for mr. many promise therefore mr. many promises will 60 percent of people will vote for him in the election and he'll win see what I mean we haven't actually looked at representativeness yet we will about to do so know you're all dying to get on to unrepresented so let's do so here we go okay this the second thing that we ask is how representative is the sample what what you should do instantly I'm giving you a again you might see it there's another algorithm another just list of steps that you might do again try and keep them separate in your mind because if you tick off each one okay you've asked yourself how many there are in the sample and how many there are in the population and made a judgment about whether there is nothing the sample to be able to extrapolate second question you asked is whether the sample is representative see what I mean didn't Descartes very famous philosopher brilliant philosopher had a list of rules of thinking and one of the things he said was that you should take any problem you have and break it up into its parts and then deal with each part separately and then make sure that looking at each of the parts you can put together as a solution to the whole and what I'm suggesting is you ask each of these questions separately so that you make sure that you ask all of them I mean the it just makes your thoughts clearer again as with first you identify the arguments and analyze it then you evaluate it okay you don't try and do both at once okay so here again you got all these where the voters sampled all female well I mean there are a lot of medical experiments or medical surveys that look only at men and then extrapolate the results to women I don't know if you've seen recently they've decided that for women the symptoms of a heart attack are quite different from the heart attack symptoms of a man and therefore all the extrapolation that they've done in the past from male experience of heart attacks to female experience of heart attacks has been faulty there was quite a big thing about that a couple of weeks ago are they all over 40 are they all white are they all middle class and they all known to the person conducting the survey the famous example that you were mentioning a minutes ago and in fact that you've just mentioned as well in an election between who Roosevelt and Landen that's right they thought that 60% of the population was going to vote that's what their sample have said but how did they find the sample they looked in the telephone book how many people had telephones then actually very few so although there was 60% of the sample said that they would vote for Roosevelt actually the sample was horrendously unrepresentative because it was middle class people with fair amount of money who had telephones and therefore it didn't represent the population as a whole okay and anyway the same thing here again we Camus have Ione wrong BT on the Sunday after 10:00 p.m. when I'm in a hurry etc etc okay so first thing is the premise true secondly what was it how large is the sample as a percentage of the population thirdly how representative is the sample three questions to our square here's another one here's a one you haven't thought of and perfectly reasonable that you shouldn't if you were asked here are two hands of cards which one is most likely to come up who thinks this one is most likely to come up nope okay who thinks this one is most likely to come up though you're all very clever aren't you you're absolutely right and they're actually equally likely to come up because of course cards are just at random they're not but actually if you if you ask the students at the University where this experiment was done which handed lights come up they come out overwhelmingly against this one and for this one this is much more likely to come up than this one now you can see why they think this can't you can you this is this one well yes this is the one they'd love to have come up and this is the one that they have come up they think all the time sort of thing but of course actually it doesn't quite work like that because they're using an informal heuristic to say in my experience this never comes up and this always comes up and actually you just can't use that here can you because what comes up is something like that but certainly not that it just means a way of making a decision okay rule of thumb if you like a way of making a decision thank you for asking I should have explained it before okay so so if an inductive generalization is based on on an informal claim like this in my experience hands like this never come up therefore this one is much less likely than that one then you you should be very wary of the generalization and here's another one and I expect you're all going to be clever enough to get this - okay in four pages of a novel how many words would you expect to find ending in Inge and in four pages of a novel how many words would you expect to find that includes the letter n would you expect that to be larger than that or vice versa so you'd expect more being words know more of the N words put up your hands if you think there are more n words okay put up your hand if you think more in words okay that's interesting this time you have fallen for the trick and because of course there you're going to be that's right they're always going to over okay I'm sorry so you're absolutely right there are going to be many more n words than there are in words because they're all they're going to be at least as many n words as there are in words yes okay sorry you did get that what happens again when you ask these students the psychology students at the University where this experiment was done is they expect many more of these because they can think of many more in words than they can of n words and therefore they they inductively generalize again well I thinking many more of those therefore there probably are more of those again bad arguments if I ask you how many footballers or something from a particular team score well you'll be able to think why I'd say that's a very bad example anything that you think you know a bit about you're probably tempted to rely on your own experience to make an inductive generalization that can work if you really do know what you're talking about but it doesn't work if you're just using that way of doing it on another context where actually your knowledge is not so secure okay okay so five steps there I think it was when you're evaluating any inductive generalization you're looking for firstly is the premise true secondly does the sample size of the population is it large enough compared to the population as a whole thirdly is the sample representative or is there a bias in it due to whatever all sorts of reasons for different biases and finally is it based on on an informal heuristic that actually an informal rule of thumb that actually just won't stand up to proper scrutiny here okay and as I said all inductive general all inductive arguments are based on inductive generalizations and so that little way of testing things can be used for all of them let's look at causal generalizations okay a causal generalization is a type of inductive generalization the premise identifies a correlation between two types of events and the conclusion states that events the first type cause events of the second type so the idea is that if you see a and B a and B a and B a and B A's and B's are always correlated you extrapolate to the claim that a z' and B's will always be correlated and you imply that the reason for this is that there's a causal relation between them so where there's correlation there's cause that's what a causal generalization is so let's have a look at a couple okay married men live longer than single men therefore being married causes you to live longer I apologize for this one when air is allowed into a wound maggots form therefore maggots in wounds are caused by air being allowed into the womb wound this is sorry okay I tell you what let's let's do it openly what do we need to know to know whether this are these arguments are good arguments okay let's have a look again we ask it to the premise true who says men married men live longer married men a woman who wants to get married Fred whose parents split up when he was five I mean who's saying this when are we actually getting this information from who says maggots formed when I get since the womb just as you said at the back there was it a newly qualified nurse who's who's observed this once was it an elderly doctor who's seen it a lot but only in his own experience and in his own study perhaps or was it a scientific study and one that you would expect to be to have looked more carefully causation is is actually I mean to give you a little bit of background on this David Hume the person I've mentioned already in connection with the principle of the uniformity of nature believes that actually causation we cannot determine causation if we find a causes B and we try and find out why a causes B what what is this causal relation what is it that relates the two things that have cause and effect we'll just find another correlation C and D okay and so why do we think C and D are correlated we look further and we look down we see yet another correlation so all we ever see is correlation we never actually see the causal relation itself we can never get to the causal relations himself itself and he actually thought arguably this is a very popular theory of human although lots people deny it these days that he actually thought causation didn't exist at all that causes that our beliefs about causation are just a habit of mind so we see a correlative would be a correlator would be a correlative would be and we start to say that a causes B and all we mean by that is that a is correlated B there's just a constant conjunction between a and B there's nothing that makes a cause B I have to say that there's another theory of Humes that he says that a causes B where has it not been the case that a it would not have been the case that B had it not been the case that a it would not have been the case that B and that suggests that there's a power of some kind isn't that that makes a cause b but but we don't ever see that power do we we don't we just see the cause and the effect and the correlation between them and so causation is a really interesting philosophical issue it's the question what causation is is endlessly interesting I think it's endlessly interesting but it remains to be the case that our evidence for causation is always a correlation but a correlation simply isn't sufficient as evidence for causation is it because it could be evidence for identity for example so that night well the evening star goes down the morning star rises and so on and so forth do they cause each other to do it no they're actually they're the same thing that's why they're correlated that's why the the pattern is uniform do husband's cause wives but they're correlated well what we're saying is that a correlation isn't sufficient for a causation but it's the only evidence we're ever likely to have but when when you see a causal generalization it will be based on correlations but what we're alerting you to here is that a correlation isn't sufficient for a causation you need to ask lots of other questions so the premise true how strong is the correlation how many married men were observed I mean this is again exactly the same as how many are in the sample from last question how long were they observed were unmarried men observed how many cases and maggots forming were observed and because what John Stuart Mill famous philosopher English philosopher came up with what he called the method of agreement and the method of difference for scientific experiments what sure if you're trying to work out what causes what's you need to see firstly that they do correlate that that the cause correlates with the effect next thing you need to do is to try and bring about the cause without the effect is if you're saying that a is cause B because all ages are always correlated with B then what you do is you try and bring about with an a without a B because if you can do that you've disproved your claim about causation see what I mean and that shows us that we tend to think that a cause is sufficient for its effect that if a causes B the occurrence of an A must be followed by the occurrence of a B because a is sufficient for B so that that's the method of of sameness --is and the method of differences which tells you whether something's are caused or not also you want to ask does the causal relation makes sense or could it be accidental let's say that we discovered that in the whole history of the universe every time a match has been struck a pineapple has fallen okay we have a correlation and we've done our very best to try and make sure that we've struck matches without a pineapple falling and it could keep on doing it some we can't break the correlation in any way do we think that matches striking cause pineapples to fall well some people are quite inductively bowled here they think yes if you got a correlation as strong as that it must be causal apparently there's also a correlation between the length of skirt and the Dow Jones index as one goes up the other goes up as one goes down the other goes down where it might be the other way around but anyway there's a correlation here do we think that the length of skirts causes the rise and fall of the Dow Jones index or vice-versa you can sort of see a something that makes sense can't you in that because you could maybe when the down don't Jones index is really high people are really excited and pleased and therefore they risk taking so they put on their miniskirt okay this does okay so the claim that's being married makes you live longer if you're a man why why would being married cause men to live longer I think this is where your claim about are we including civil relationships is quite interesting why would being married cause men to live long instantly I think it causes women died earlier just a warning to women in this room okay they're happier they their stress reduces another explanation what that's the first decent it might also be because women tend to look after diets and things like that more men are cooked for more often than women are perhaps and when women do the cooking they concerned about nutrition and da-da-da-da-da so when a married man eats he tends to eat more healthily in honor I mean we can think of reasons for why that would be the case can't we so it's not a complete mystery what about this one why would air getting into a wound cause maggots to form so I'm in the experiment we've done here we some nurse has seen that when a wound was covered up by accident or something like that maggots didn't form and she thinks well we know could it be so she covers up a few as she leaves a few open and she sees that the one she's covered up don't get maggots whereas the ones that left open do get maggots so she's formed a hypothesis could it be that air getting into the wound causes maggots but why would that be the case perhaps because there's something carried in the air that causes maggots to form and actually we know now that that is the case so okay so does the course of relation make sense incidentally if it doesn't make sense does that mean it's not causal no it doesn't actually does it's QC you can imagine that there may be something that is a complete mystery for us for a while and I wish I could think of an example but which turns out to be true and turns out to have an explanation but even so if you can't if it just if the things seem to be just totally disparate that would be a mark against this argument being a good one and we also might and we've we've done this a bit okay what's caused it is what could it be that being long-lived causes marriage so it might be that having genes for longevity cause men to get married so you said socioeconomic factors but I'm suggesting it could be genetic factors so there's one set of genes such that if a man has them he's both more likely to get married and he's more likely to live longer so there's one common cause for the two things rather than that one thing causes the other yeah and that's I couldn't think of anything good maggots forming cause yeah they're to get into the womb no I couldn't think about that so okay right that and so that's looking at causal generalizations and you'll see that many of the questions that you would ask about causal Jen right generalizations are also questions you've already asked about inductive generalizations that's not surprising because causal generalizations are a type of inductive generalization and all the ones that you're asking separately the ones that say you know why should we think that a correlation has a causal relation under it so that so just moving on quickly to analogy here another type of inductive generalization it takes just one sample of something and then extrapolates from a character of Latics or to the character of something similar to that thing and there's a famous argument from analogy the universe is like a pocket watch pocket watches have designers therefore the universe must have a designer and I think we're probably all familiar with that arguments okay how would we go about questioning this argument yes okay what aspects are we picking out here and saying is similar to the two cases so why is the universe like a pocket watch I mean using this famous example what did the person believe that's right it was very who was it it's gone completely Paley that's right thank you I'm sorry I have got a head full of cotton wool it's very strange yeah Paley believed at the universe the pocket watches is moves regularly it's very complex it's it must have been very difficult to put together and he believes that the universe is also very regular very complex it must have been difficult to put together therefore if one has a designer the other husband's designer what else might you ask okay there are many there is a similarity we might say between pocket watches and the universe but there are many many dissimilarities why should we consider that this similarity is more important than all these differences yep okay but wouldn't you say that if the universe is like a pocket watch in this particular thing and the explanation of pocket watch is having a designer is this particular thing in other words it's being very complex so if we agree that everything that's complex and regular must have a designer okay but we are saying the the universe is like a pocket watch in being very complex and regular pocket watches have a designer oh I see okay so you're absolutely right I'm sorry no you are right I was I was changing that second premise to everything that's complicated has a designer and that's not what it says is it and so I've rightly been pulled up on that okay it isn't what it says I suppose that's why we think that this is going to work at all though isn't it in order to give an argument we do have to say a lot of things in support of the various premises and in support of our belief that the conclusion follows from the premise and so on so you wouldn't expect almost anything said to be in arguments and actually as you learn yes no I'm not surprised it I'm Jose what I'm doing is undefended newspapers because actually you need to read a whole whole article in order to see what the claim being made is and then you need to go back and identify what the reasons are being given for the claim okay are the two things similar in theirs and the respective is there is the respect in which they're similar relevant to the argument being made and also can we find a dis analogy which is the thing you mentioned is are there differences between them and do the differences pertain to this arguments but the thing to remember about arguments from analogy is that they are extrapolating from just one example therefore the one example and the extrapolation have to be really pretty strong before you should go along with them so arguments from now actually arguments through analogy a much more common and probably for the reason you're saying because they they often take us along with them emotionally let's finally look at arguments from Authority which take one person or a group of persons who are or are assumed to be right about some things and they extrapolate to the claim that they're right about other things so human rights monitoring organizations are experts on whether human rights have been violated they they say that some prisoners are mistreated innate in Mexico therefore some prisoners are mistreated in Mexico what do we need to ask about this where do they get their information from is it just that they've become hackneyed and cynical and they think that everyone miss treats everyone or do they actually have reasons for saying what they have yep I mean all that's needed for this argument is that some prisoners are mistreated not that they're mistreated by anyone in particular I think okay you might say here we've started the first premise here is they may be experts on whether human rights have been violated but are they experts on whether somebody's being mistreated and are they perhaps seeing trivial forms of mistreatment as violations of human rights or something is that what you mean yep okay okay well let's have a look at the okay who exactly is the source of information I mean it was saying there it was implying at least that all the human rights organizations were saying it but it might just say one and again there you would make you want to make a judgment about whether the source of information really is an expert whether they're qualified in the appropriate area because it's very easy again going back to how inductively bold you are if you have a tendency to think this person is an expert in one area you may well inductively generalize to his or her being an expert in another area so your tutor for example whom you think is you know if Maryanne says P then P which is of course a very good argument but if what she's talking about is politics or mathematics or something like that then it's complete nonsense isn't it okay so so not only that you need to know who they are you need to know whether they're qualified in the right area you need to know whether they're impartial in this in respect of this particular claim so Amnesty International let's say our our impartial they go out and they get the evidence and they're very careful not to be biased I don't know whether that's true instantly but I let's say it could be but then there might be another human rights organization that's not careful to make sure that it's information isn't biased so you'd need to make a distinction between the fact that amnesty is a reputable organization and this other one isn't well I mean you get that quite often I mean I mean if you want to belittle the results that come out of a particular survey one way of doing it is to say that the people who are who are putting forward this survey are biased so I've been working on looking at GM food and actually it's very very difficult to to get a source that hasn't been funded by a pharmaceutical company or by a company that isn't that that's the sort of the Soil Association or somebody that's very anti GM food so finding something that really is an impartial source is really very difficult and it's very very important to try and find one if you're really going to evaluate these arguments finally the the points you made a minute ago it's very rarely the case that you have one expert in an area and it's very rarely the case also that all the experts in an area will agree on on something and if you have different experts making different claims you need to make a judgement as to where you think which of them do you think is it's correct and what you can't rely on there is is an argument from Authority can you because they're both authorities so if you were an undergraduate writing an essay or indeed if you were you writing an essay on philosophy for me I would have given you lots of reading you would have done the reading and I would have expected you to come away and to think okay well so-and-so says this and finger-mabob says that and he says P and he says not P well which of them is the case well now you need to look at what the arguments are that so-and-so gives what the arguments are that thingamabob gives and work out which ones you think are the best ones and why okay so there's no substitute for thinking for yourself and an appeal to an argument for authority is okay for for various things I mean we have to rely on authorities for all sorts of things but if you were trying to write a philosophy yes saying Maryann says P therefore P will not do and that's true of every philosopher you ever come across because there are very very few things in philosophy that aren't questioned ok that's where I was going today next week we'll look at validity and truth and then we'll turn to the evaluation of deductive arguments

Channel: University of Oxford

Views: 20,357

Rating: 4.9375 out of 5

Keywords: yt:stretch=16:9, philosophy, arguments, critical reasoning, argument, reasoning

Id: UfIhvMhno2w

Length: 53min 9sec (3189 seconds)

Published: Fri Oct 22 2010