Social Arguments

16 Testimony and Trust

Section 1: Introduction

Earlier we noted that other people are probably the single most important influence on our thinking.  We learn a lot about the world and how it works from others, and in this unit we will focus on this process.  More specifically, we will focus on arguments that draw a conclusion about what to believe on the basis of what other people (either individually or in groups) say or are like.  As we will see, when it comes to learning from others perhaps the most important step involves judgments of credibility, and we will look at how we make these judgments and what makes a judgment of credibility logically strong.  In addition, we often check with others about our beliefs, and use their agreement or disagreement with us as evidence for or against the accuracy of those beliefs.  Finally, we will look at what groups can teach us, and under what conditions.  After all, other people are rational and make good choices, and when they are doing so, we can use their actions and beliefs as a guide for our own.  Of course, this is not always the case, and in Chapter 18 we will distinguish the two.  However, in this chapter we will begin with learning from what people tell us, that is, with testimony.

Section 2: Testimony and Trust

If you stop and think about it, a lot of what you believe, you believe because somebody has said so.  For example, you might believe that your mom had a cavity last time she went to the dentist because that is what she told you; you might believe that x-rays were discovered in 1895 because that is what the textbook says; you might believe that Slumdog Millionaire won the Oscar for Best Picture in 2009 because that’s what came up when you did an internet search.  These are all cases of believing something on the basis of testimony.  As we will use the terms,

A person believes something on the basis of testimony when (and only when) they believe it primarily because somebody else claims it is true.

The term ‘testimony’ is probably most familiar to us from legal settings.  In a trial, for example, a witness might be asked to give her testimony to the jury.  In this case, the jury is being asked to believe that what the witness says is true.  Testimony, as we will understand it, need not take place in a courtroom or legal context, however.  The definition above is wide and includes any case a person comes to believe something because somebody else claims it is so.  This includes believing things that other people have written down, as well as things they say.  To believe that your roommate will be back at 5 pm because that is what the note says or that Congress passed a bill because that is what the newspaper reports are both cases of believing on the basis of testimony.

As the examples above suggest, we rely on the testimony of other people all the time—and for good reason.  After all, the knowledge and experience of a single individual is quite limited.  Think about how you know about something as basic as the circumstances of your own birth, for example.  You can’t remember it, and although there might be pictures or videos of you as a baby, you know many of the details only because your parents or relatives have told you about them.  Testimony allows us to easily extend our knowledge far beyond the limits of our own inquiry and experience.  In this way, you do not have to go with your mom to the dentist to know about her cavity or do the research to find out the history of x-rays—you can simply take others’ word for it.  Indeed, it is difficult to overstate the importance of testimony, and hopefully it is clear that our understanding of ourselves and the world is due, in large part, to our wider community.

Importantly, to accept somebody’s testimony is to trust that they are a reliable source.  When we trust others, we open ourselves to new information, but we also open ourselves up to deception and exploitation.  After all, we know that sometimes people lie to us or otherwise try to deceive us.  For example, imagine your phone rings and the person on the other side identifies himself as a representative from Microsoft’s technical support.  He goes on to tell you that your computer is infected with one or more viruses or pieces of malware, and that he will need remote access to your computer to fix it.  Put in these terms, this is obviously a bad idea.  It is important to remember, however, that smart people fall for these kinds of scams all the time.  Presumably they do so for all kinds of reasons, but anybody who succumbs to this scam does so, at least in part, because they have trusted the stranger on the other end of the phone call—they have accepted this person’s claim that they are, in fact, who they say they are.

Map of Michigan showing places where there have been numerous phone scams reported
“Consumers Energy warns of phone scams” by Consumers Energy is licensed under CC BY-NC-ND 2.0

Ideally, then, we would trust only reliable sources, and would be able to distinguish reliable from unreliable ones.  As it turns out, our normal approach towards the testimony of other people aims toward this goal.  Cognitive scientist Dan Sperber and others call our natural approach to the testimony of other people, vigilant trust: we tend to accept what other people say as true while at the same time keeping an eye out for signs that they are not trustworthy.[1]  Thus, we seek to learn as much as we can from the testimony of other people without being gullible or naïve. Moreover, this is an effective strategy, and in many cases people are both sincere and accurate in what they say.  There are plenty of exceptions, of course.  Everybody has lied to or misled others, and sometimes even our sincere beliefs turn out to be false.  Thus, an attitude of vigilant trust roughly corresponds to the world as we find it, and is an expedient approach to navigating our social situation.  Vigilance takes different forms.  On the one hand, much of our vigilance is automatic.  As people talk, we implicitly monitor their tone, body language, and so on for cues of insincerity or incompetence.  On the other hand, we can be vigilant much more explicitly by actively looking for relevant characteristics of the speaker or source.  Before we take a closer look at vigilant trust, it is important to make a couple of brief notes about the evidential value of testimony.

Testimony is normally a quick and easy way to add to our system of beliefs.  Nonetheless, it is important to be aware of the limits of this kind of evidence.  First, testimony is not particularly weighty evidence.  Although, it can certainly give us good enough reason to believe in some cases, in many cases it does not—even when the source is credible.  Imagine, for example, that your friend tells you that your professor has moved the final exam to a different time.  You may trust your friend, but it is always possible they have made a mistake—and if so, you’ll miss the final, and miss your chance for a good grade as well.  Since the consequences of being wrong are substantial, you’ll want to check on this for yourself instead of simply taking your friend’s word for it.  More broadly, in cases where the claim in question really matters or has significant potential costs, you will want to look for additional evidence rather than relying solely or primarily on testimony.

Second, compare the following two cases: (a): you work a math problem and come to the conclusion that the answer is ‘42’; (b): you don’t work the math problem, but come to believe that the answer to it is ‘42’ because your friend tells you the answer.  Let’s say that the answer really is ‘42’, and so you’ll have a true belief either way.  What is the difference?  One big difference is that in (a) you will understand why the answer is ‘42’ and in (b) you won’t.  We can generalize on this point to say that, at least in many cases, testimony does not help us understand why something is true, only that it is true.  Of course, in many cases we do not need to have a deeper understanding; knowing that it is true is good enough.  Moreover, testimony can provide information that will allow us to explain or understand broader issues.  What a witness tells the jury can help it figure out who committed the crime, for example.  Nevertheless, understanding something for yourself, and knowing because somebody told you are very different ways of knowing.  In at least some cases, it is important to understand why something is true, and in these cases testimony will not suffice.

Section 3: Testimony and The Credibility Assumption

As we saw in previous chapters, there are largely automatic and implicit reasoning processes that can inform our conscious judgments, and trust is no different.  As we navigate interactions with others, we are constantly and implicitly monitoring them for signs of dishonesty or incompetence.  Thus features of the situation or source can leave us with a vague sense that they are trustworthy or suspicious or something in between.  While it is important to be aware of these automatic processes, we will focus in this chapter on conscious and explicit inferences.  What is involved in choosing to be vigilant?  Let’s start with a simple case.  Say your friend Sierra tells you that your English class is cancelled for today, and on this basis you do not go to class.  We can represent your reasoning this way:

  1. Sierra says class is cancelled for today.
  2. So, class is cancelled for today.

Formulated in this way, we can see there is a gap in this argument.  What connects the premise that Sierra says it with the conclusion that it is true?  In general terms, we might say—well…Sierra wouldn’t tell me that it is cancelled if she wasn’t sure about it.  That is, because Sierra is a trustworthy, reliable, or credible source on this issue.  More specifically, we will say that a source is credible on some issue only if:

(i) They are saying what they believe to be true, and

(ii) They are in a good position to have an accurate belief.

In light of this, we will call the assumption that connects a person’s claim that something is true, to the conclusion that it probably is, the Credibility Assumption.

Because (i) and (ii) are requirements for credibility, if we have reason to suspect that either (i) or (ii) is not the case, then we have good reason to doubt the source’s credibility. That is, we are justified in doubting a person’s credibility on some issue if we have reason to think they are not saying what they believe to be true, or if we have reason to believe that thy are not in a position to have an accurate belief on the issue in question.  To be clear, to doubt a person’s credibility on some issue is not to conclude that they are lying or saying something false.  Rather, to doubt a person’s credibility on some issue implies only that their word is not, by itself, good enough reason for you to believe what they are claiming; you will need additional evidence before you will accept what they’ve claimed.

Judge's Gavel
“Court Gavel – Judge’s Gavel – Courtroom” by weiss_paarz CC BY-SA 2.0

So, what might give us reason to doubt The Credibility Assumption?  The most natural place to start is with the source itself.  Indeed, there are a number of characteristics a source might have which would undermine its credibility.  The most obvious example are people who have a history of regularly lying, deceiving, or exaggerating.  The known perjurer, for example, is not somebody to whom we should give the benefit of the doubt.  There are a variety of other characteristics that give us good reason to doubt a source’s testimony.  In order to get these on the table, consider the following (fictional) example:

Max is on trial for the murder of Ethan.  While on the witness stand, Isabel claims that she saw Max leaving Ethan’s house around the time of the murder.

Let us suppose that Isabel does not have a history of regularly lying, deceiving, or exaggerating.  What other kinds of information might undermine Isabel’s credibility?  Imagine that we discover the following facts about Isabel and Max’s relationship.

Isabel is Max’s ex-wife, and the split was not amicable.  While the two were married Max gambled away all of the couple’s wealth, including a substantial inheritance Isabel had received upon the death of her parents.  Moreover, Max repeatedly cheated on Isabel during their marriage, and Isabel is now aware of this fact.  As a result of this, Isabel blames Max for ruining her life.

This sad story undermines the value of Isabel’s testimony, because it suggests a possible motive for lying.  After all, it is possible, given what we know about human nature, that Isabel might lie in order to get revenge on Max for ruining her life.  Put otherwise, it is possible that Isabel has an interest in Max’s being convicted of the crime—regardless of whether he actually committed the crime or not.  Whether a person has an interest or something to gain is perhaps the single most important factor in assessing a person’s credibility.  In general, the more a person has to gain, the less credible they are, and correspondingly, the less they have to gain the more credible.  There is a deeper point here, as well.  In this example, the fact that Max has ruined Isabel’s life gives her an interest or motive for lying, but as we saw in Chapter 9, interests can influence us in less explicit ways.  Recall that simply have a preference for one claim or outcome over another can bias our evaluation of information.  Thus, knowing that a person has an interest in some claim gives us some reason to think they might be biased in their thinking, and consequently, some reason to be suspicious of their testimony.

Of course, in spite of all these factors Isabel might be telling the truth—she might have seen Max leave the house.  Nonetheless, in light of these facts, Isabel’s testimony has much less weight than it would have otherwise; the jury cannot merely take her word for it and accept her testimony at face value once they know the background.  Let us put aside the story that Max and Isabel were married, etc., and consider an alternative.

Suppose that Isabel was 50 feet away when she claims to have seen Max leaving the crime scene.  Furthermore, it comes out during cross-examination that Isabel normally wears glasses because she is extremely near-sighted (she has an uncorrected visual acuity of only 20/200), and she was not wearing her glasses that day.

Again, this information does not necessarily offer evidence that Isabel is lying, but it does undercut her credibility.  The fact that her uncorrected vision is so poor suggests that she was not capable of seeing accurately enough to identify Max as the person leaving the crime scene.  Thus, even though she may believe what she is saying, she is making a claim that is beyond what she could reasonably know, and thereby gives us reason to doubt that her claim is correct.

There is one other related factor it is important to note.  Consider the following:

Suppose that Isabel did not see the murderer’s face, but did get a look at the handgun used in the crime.  Isabel specifies that the gun used in the crime was a 9mm Beretta PX4 Sub-Compact—exactly the same kind of handgun owned by Max.  It comes out during cross-examination, however, that Isabel is unable to distinguish between the Beretta Compact and Sub-compact models, and moreover is not even able to distinguish between different brands of 9mm handguns.

That Isabel cannot distinguish these three kinds of 9mm handgun strongly suggests that she does not have enough background knowledge about handguns to reasonably know that the gun she saw was a 9mm Beretta PX4 Sub-Compact.  This is not something she could reasonably know because she does not have the knowledge base or expertise to make this distinction.

We have considered four especially common factors to be on the lookout for in considering a person’s credibility on a specific issue.  There are a variety of additional complications that we will take up in the next chapter, but for now we will turn to a second factor to keep in mind in deciding whether to believe somebody’s testimony.    

Section 4: Checking for Plausibility

When we are being vigilant, we need to check for credibility, but we also need to check the claim against our general knowledge of the world (as per the The Rule of Total Evidence).  That is, we need to ask—is the claim plausible given what we know?  To illustrate, imagine that a friend tells you they saw former President Bill Clinton buying a toothbrush at the local pharmacy.  Your friend seems sincere and there is no reason to think she is deceiving you.  Moreover, she certainly knows what Bill Clinton looks like.  In short, there is no reason not to trust her.  The problem is that, as you think about it, it just seems so unlikely.  You know that Bill Clinton doesn’t live in the area, and what are the chances he would be buying a toothbrush here?  Put in different terms, this claim seems implausible given your knowledge and experience.  You are thus left with competing pieces of evidence, and you need to weigh them against one another.    Which is more likely: that Bill Clinton really was at the local pharmacy buying a toothbrush or that your friend has misidentified somebody who looks a lot like Bill Clinton?

The main point here is that credibility is only one factor and can be outweighed by the implausibility of the claim.  Overall, then, in being vigilant we’ll want to ask two main questions.

Two Questions to Ask of Testimony

  • Is the source credible? (history of lying? something to gain? capability/expertise?)
  • Is the claim plausible given what you know about the world?

Again, to set aside what somebody has said because you find them less than fully credible or because you find their claim implausible is not to disagree with them or to conclude that they are wrong.  To set their claim aside is simply to say that testimony is not enough in this case, and that more evidence is required before you’ll be willing to add it to your system of belief.

Section 5: Testimony and the Web

We are constantly turning to the web for information.  We can, for example, type a question into a search engine, and get a wide variety of answers almost immediately.  Even when we are not actively searching for information, we are continually presented with news, commentary, opinions, humor, memes, factoids, and so on as we scroll through our friends and contacts on social media. This observation is relevant because it highlights that the internet is what philosopher Michael Lynch calls a testimony machine.[2]  Think about it: the information on the internet was originally posted by somebody, and so when we come to believe something because a website, meme, or tweet says so, we come to believe it on the basis of testimony.  Indeed, the web vastly increases our access to information in large part by vastly increasing our access to other people’s knowledge and experience.  At the same time, however, it also increases our exposure to false, misleading, and deceptive sources.  For example, in recent years headlines shared through social media have proclaimed that the Pope endorsed Donald Trump for President, Minnesota made Arabic classes mandatory in high school, and the U.S. government issued warnings about bananas contaminated by the Zika virus.  These are all false, but were widely shared on social media.  As another example, according to a study performed in the Spring of 2021, close to 65% of all anti-Covid vaccination content shared on Facebook and Twitter could be traced back to just 12 people![3]  Thus, while the web greatly expands our access to information, it simultaneously allows false and misleading claims to be quickly spread and amplified.

A social network arranged by color
“A social network visualization” by brewbooks CC BY-SA 2.0

A big part of the reason that people are fooled when it comes to internet sources is that “web environments” make it difficult to be vigilant.  As we’ve seen, an important part of being vigilant is looking to the credibility of the source, but the web presents us with many unfamiliar sources.  Indeed, in many cases the source’s identity is hidden.  That is, online we are often not in a position to see whether there are factors undermining a source’s credibility at all.  Further, even when we are familiar with a source, it can be difficult to pick up on signs of dishonesty, insincerity, or incompetence on the web.  Thus, in many cases online we are flying, at least partially, blind.

There are two online contexts it is worth emphasizing in particular on this issue: using a search engine and using social media.  A search engine is a handy tool for navigating the sea of information available on the web, and often the first thing we do when we have a question is to “Google it”.  In thinking about web searches it is important to distinguish at least two kinds of question you might use a search engine to answer.  On the one hand, you might have a question about a specific, straightforward, and uncontested fact.  For example, we might need to know how many cups are in a gallon or the capital of Peru.  We can trust search engines to quickly deliver accurate answers to these kinds of questions.  Often, however, we are interested in questions that are more evaluative, interpretive, complex, or disputed.  Indeed, most of the important questions we have fall into this second category, and when we use search engines to look for answers we need to be careful.  Why?

When we type our question into a search engine, we get a ranked list of links.  Some of these links may be to trusted and familiar sources.  However, most of the links will be to sites we’ve never heard of.  This means that we won’t know anything about who is behind the information posted on these sites, why they posted it, and whether they are in a position to know what they are claiming (without extra work).  In short, we won’t be in a position to say anything about the credibility of most of these sites.  This is not a criticism of search engines as much as it is reminder that internet searches quickly and easily take us to “strangers” whose credibility we may not be able to assess (at least immediately), and whose testimony we should therefore be cautious about.

Man holding up mirror in front of face so his identity is obscured
“Hidden Fire.” by Ardinnnn 🙂 CC BY-NC-SA 2.0

But wait.  The search engine’s algorithms have identified this particular ranked list of sites in response to your question.  Doesn’t that make these results trustworthy?  No.  While it is true that a search engine brings potentially relevant sites to your attention, search engines do not work by identifying only trustworthy sources.  Instead they work, in very general terms, by looking for key words, and finding sites with those key words that are popular (among other factors).  Indeed, when you do a search, we end up with all kinds of results.  You might get links directing you to online forums like Reddit or Quora, news sites, blogs, Wikipedia or other online encyclopedias, think tanks, trade magazines, videos, advocacy organizations, academic sites, and online businesses to name just a few.  And these sources will not be equally reliable or trustworthy sources for answering your question.  There is a deeper point here as well, namely that while it is true that a very reliable source might be popular as a result of its reliability, the reverse doesn’t hold—the popularity of a site doesn’t tell us a lot about its reliability, since there are many reasons a site might be popular besides it’s being a reliable source of information as we will see in Chapter 18 (it is entertaining; it tells people what they want to hear, and so on).  Consequently, we can’t take the mere fact that a link came up in response to our question to mean that it is a trustworthy or reliable answer.

Turning to social media, we all know that you can’t trust everything you see on Facebook (for example).  Nevertheless, people are regularly fooled by headlines, stories, and pictures that are shared by their connections.  What accounts for this?  In some cases, we know the source of information that is shared, for example when a friend shares a news story from a well-known and trusted site.  There are many cases, however, where the origins of the shared information are unknown.  Again, in these situations we are not in a position to see whether the source is credible or not.  This is made worse on social media by a number of other factors.  First, think about your normal state of mind when you turn to social media. We go to social media for entertainment and to see what our friends and acquaintances are up to, and as we causally scroll through our feed we tend to be in a relaxed and passive state of mind.  Second, the way social media feeds look can lend the claim of an anonymous, hidden, or unknown source an undeserved air of credibility.  After all, the headline, story, or picture was posted by your friend, and is situated right under their name and picture.  It can be easy, especially if you aren’t being too careful, to extend the credibility of your friend to the claims they’ve shared.  Third, sometimes people simply don’t care whether what they share is true.  A person might share a headline, story, blog post, or picture on social media primarily because they find it entertaining or interesting.  Alternatively, they might do so primarily to express their identity or membership in some group.  As we’ve seen, ideas can have social value, and just as this fact can lead us to have preferences for or against certain ideas, so too can it lead us to share things without a particular interest in their truth.

We’ve highlighted situations in which we are especially likely to run into unfamiliar sources of information.  As we’ve seen, in these cases we simply can’t say anything about the credibility the source, and we should be careful.  This isn’t to say that we can never trust a source we don’t know much about.  If an online source tells you something simple, objective, undisputed, and about which there is no obvious reason to mislead, then it might make sense to take its testimony (e.g. How many feet in a mile?  What planet is closest to the sun? What is the official currency of Brazil?).  However, most interesting claims on the web are not like this.  If a claim is surprising, too good to be true, or the sort of thing a person might lie about, then you won’t want to take a stranger’s word for it.  Again, this is not to conclude the claim in question is false or that the source is trying to mislead you, only that it is going to take more than the source’s word to convince you the claim is true.

Section 6: Summarizing Testimony

There are three points to emphasize in summary.  First, vigilant trust aims to steer between being naïve or overly trusting on the one hand, and being overly skeptical on the other.  It is to strive for the informational benefits of reliable sources without being taken-in or otherwise misled by unreliable ones.  Thus, we want to be open to learning from the knowledge and experience of others, while simultaneously keeping an eye on their credibility and the overall plausibility of their claim.

Second, as we’ve seen testimony is an important source of information, but it has limitations.  In particular, although a person’s testimony can give us good enough reason to believe it, testimony does not help us understand.  That is, although testimony can give us good reason to think that something is true, it doesn’t explain why it is true. If the claim in question becomes contested or has significant practical consequences, you’ll want a more comprehensive understanding than testimony can normally provide, and you’ll want to look for further evidence.

Finally, we have been talking about how to navigate people’s claims on social media and elsewhere, but this also opens the door to thinking about our own behavior and the kind of reputation we’d like to cultivate.  Thus, we can turn things around and ask: do other people see us as a credible source?  Are we being sincere in what we say, post, and share?  Do we share information we are in a position to know?  Or do we suggest things are true, when we really aren’t sure?  More broadly, it puts us in a position to think about our responsibilities to others when it comes to testimony.  This is a particularly pressing question when it comes to social media, given its ability to spread and amplify information.  What do we owe to our friends, followers, and connections on social media when it comes to information?


Exercise Set 16A:


List three things you believe on the basis of testimony.


What are the last three things you shared on social media?

Exercise Set 16B:

Directions: Let’s practice being vigilant.  In each case, can we take their word for it?  Why or why not?


Mr. Ziegler has testified that his son Scott was home with him at the time when Scott is alleged to have shot the victim, so Scott can’t have been the shooter.


“This place has the best Chinese food in Chicago” says your friend who just moved there from out of state last month.


CEO of Massive Dynamics (a company that makes vitamins among other things):  There is no need to be concerned; our vitamin supplements are perfectly healthy.


Used car saleswoman: this car is a peach; the previous owner was a little old lady who only drove it to church on Sundays.


You are on a campus visit, and are looking for the admissions office.  You walk up to a young person walking down the sidewalk, ask, and they say, “the admissions office is on the second floor of that building right there” as he points.


“There is no way introducing this little frog into the ecosystem will disrupt it,” says the local fire chief.


The dentist says that the tooth is compromised and has to be removed.  I guess I better make an appointment to get it extracted.


Watch out for Bigfoot on your camping trip this weekend.  My friend says she saw one when she was hiking there last summer.


Dr. Johnson says that man-made global warming is a myth.  He has a Ph.D. in geology—he should know.

Exercise Set 16C:


Think back to the discussion of cooperative dialogue and cooperative disagreement in Chapter 2.  What role does trust play in these forms communication?


Suppose that a person claims to have been abducted by aliens, but is unable to offer any evidence other than their word.  What would it take for you to legitimately take this person’s word for it—that is, what would you need to know about this person to believe their testimony?


Look at your social media feed, and find something that strikes you as untrustworthy.  What is it, and why does it strike you as untrustworthy.  Be specific.


How can our actions online have positive or negative effects on others?  Think of some specific cases.  What does this say about our responsibilities to others online or on social media?



  1. Sperber, D. et al. (2010) “Epistemic Vigilance.” Mind and Language 25:4, 359-393.
  2. Lynch, Michael Patrick. (2016). The Internet of Us. New York: Liveright, 24-25.
  3. Center for Countering Digital Hate (2021, Mar. 24).  "The Disinformation Dozen.


Icon for the Creative Commons Attribution-NonCommercial 4.0 International License

Arguments in Context by Thaddeus Robinson is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, except where otherwise noted.

Share This Book