Back to the on-screen lesson ·
identify a reason offered to support a belief
Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.
You will identify a reason offered to support a belief, explaining the inference and its limits in a supplied case.
A belief can be true or false independently of being sincerely accepted. Now separate its support from its truth.
| Term | What it means |
|---|---|
| Evidence | Information offered in support of a claim. |
| Justification | Adequate support for believing a proposition under the relevant account. |
| Rebutting defeater | Evidence supporting the opposite of the conclusion. |
| Undercutting defeater | Information weakening a supporting connection without itself establishing the opposite. |
If asked why a jar contains salt, a person might reply, 'Because I believe it contains salt.' This repeats the attitude whose support was requested. A different reply would identify a label checked against the inventory, a witnessed filling, or a suitable test. These are candidate reasons because they connect the claim to something beyond its mere acceptance. Whether a reason succeeds depends on its relevance, credibility, and connection to the conclusion. Listing reasons is therefore the beginning of evaluation, not its completion.
Justification concerns the support for believing a proposition. Truth concerns whether the proposition holds. An investigator can have good evidence that later turns out to mislead, and a guess can accidentally be correct. The distinctions are easiest to see when we keep two records: what was available to the investigator, and what the case tells us was actually happening. Do not use information revealed only at the end of a story to accuse the earlier investigator of ignoring evidence they did not possess.
Suppose a museum guide consults an apparently current schedule, sees that the gallery closes at six, and tells a visitor so. Unknown to the guide, an emergency has changed closing time to five. The schedule gave some support for the guide's belief, yet the belief was false. Whether it was fully justified depends on details such as update procedures and available warnings. We need not settle every theory of justification to see that evidential support and actual truth are distinct dimensions. That is the limited conclusion this example establishes.
In this lesson, evaluate support under explicitly supplied evidential standards. The exercises sometimes stipulate that a reading is calibrated or a witness directly observed an event. Treat those as premises rather than inventing hidden faults. When the exercise introduces a known defect, include it. This makes the assessment fair: you are reconstructing how a reason functions in the given case, not declaring your personal threshold for trust or discovering an unspecified trick.
Another way: From evidence to a bounded conclusion
A reason must reach the conclusion actually offered. Seeing a light in one room supports that some light is on in the building; it does not establish that every room is occupied. A thermometer reading at noon supports a temperature claim at noon under the device's stated accuracy; it does not directly establish the temperature throughout the week. When evidence seems impressive but an inference feels too large, write down its time, location, sample, and uncertainty. The unsupported expansion often becomes visible immediately.
Separate relevance from credibility. A highly reliable attendance register may have little relevance to whether a bridge can bear a certain load. A relevant load test may nevertheless be poorly performed. Good reasoning requires both an appropriate connection and adequate grounds for accepting the supporting claim. Neither the prestige of a document nor its topical resemblance to the conclusion settles both questions. Ask first what difference the evidence would make if accurate, then ask why you should regard it as accurate.
A claim can have multiple supporting routes. A visitor can see a train depart, hear the station announcement, and inspect an updated departure board. These routes may strengthen the case, but only if we examine their dependence. If the announcement and board are both generated by the same faulty feed, agreement between them gives less additional support than it initially appears. The testimony lesson develops this issue. Here it illustrates why a count of reasons is not automatically a measure of justification: three versions of one observation are not three independent observations.
Conclusions should preserve the force of their reasons. If the case says that a sensor is usually accurate, write that its reading supports the claim, not that it proves the claim with certainty. If the premise deductively entails a conclusion, identify that stronger relation. The phrase 'has evidence for' leaves room for error; 'entails' rules out a true-premise, false-conclusion case. Using these expressions carefully lets an argument be both informative and appropriately cautious. It prevents a modest observation from turning into an unlimited guarantee.
Another way: Perception and the possibility of correction
Perception is a major source of evidence. You ordinarily learn that a cup is on a table by looking, without first constructing an elaborate proof of the reliability of vision. Yet familiar conditions can undermine a perceptual judgment: poor lighting, an obstructed view, a reflection, or a known stage illusion. Philosophers disagree about the ultimate account of perceptual justification, but such cases give a shared practical distinction between having an experience and drawing a particular conclusion from it.
Take a straight straw viewed through water. It can look bent at the surface. The experience is real as an experience; what requires correction is the conclusion that the straw itself is bent. Moving the straw, viewing it from another angle, and examining it outside the water provide checks. Describing these checks does not mean that perception is worthless because perception helped correct perception. It shows that evidential methods can be mutually supporting while remaining sensitive to identifiable failure conditions.
Distinguish a report of appearance from a report about the object. 'The fabric looks red under this lamp' may be accurate even if 'the fabric is red under standard daylight' is false. The second claim needs an assumption connecting the lighting conditions to stable color classification. If the lamp is known to be red, that connection weakens. We should state the assumption explicitly, because it is the point a critic can test. Without it, the move from experience to object seems either infallible or inexplicable when it is neither.
This is a useful place to introduce defeaters. A rebutting defeater supports the opposite of the target conclusion. An undercutting defeater challenges the evidential route without directly establishing the opposite conclusion. A verified photograph of an empty parking space can rebut the claim that a car is there at that time. Learning that your camera feed is frozen undercuts the feed as evidence; it does not by itself show the car is absent. The distinction is about what the new information does, not how alarming it sounds.
Another way: Evaluating and revising an account
Begin an analysis by stating p, the belief under review. Next state e, the evidence offered for p. Then articulate the connecting assumption: why would e make p reasonable to accept? Finally look for supplied information that defeats e or its connection to p. This four-part reconstruction can reveal circular support. If the only reason for trusting a meter is that its current reading must be correct, while the only reason for accepting the reading is trust in the meter, no independent support has been supplied for that disputed reading.
Not every chain of reasons must be extended forever in a classroom exercise. Foundationalist approaches allow some justified beliefs to receive support without depending on other beliefs in the same way. Coherentist approaches emphasize support within a suitably connected system. Externalist approaches can give a central role to how a belief was produced even when the subject cannot articulate all of that process's credentials. These are families of positions, not a single settled answer. Our exercises ask you to identify the offered support and its weaknesses before comparing such theories.
Another useful distinction is between having evidence available and basing a belief on it. A laboratory worker might possess excellent test results but form a particular belief by reading a horoscope instead. The available evidence could support the proposition, yet the actual route by which the worker arrived at it raises another question. Philosophers sometimes distinguish propositional justification from doxastic justification to track this difference. At this stage, express it plainly: what could support the claim, and what actually guided the believer?
When a defeater appears, revise only as far as it warrants. A blurred photograph might require suspending judgment about a visitor's identity, rather than concluding that the visitor was definitely someone else. A corrected timestamp might move an event to a different hour without disproving that the event occurred. The size of a revision should match the part of the argument affected. This discipline avoids two opposite errors: clinging to a conclusion after its support fails and discarding every related belief because one premise needs correction.
Your final answer should identify a reason, explain its connection, and state a limit. For example: 'The current inventory supports the claim that the jar contains salt because it records this jar's filling, but a broken chain of custody would weaken that support.' This is stronger than merely writing 'justified' because another reader can inspect the inference. It is also more useful than promising certainty, since it identifies what evidence would make a responsible investigator revise the judgment.
A school greenhouse uses a display to monitor temperature. At ten o'clock it reads twenty degrees. The technician initially accepts that the greenhouse is near twenty degrees because the sensor was checked against a reference instrument the previous day. That calibration history supplies a reason to connect the display with the temperature. The conclusion remains bounded by the device's accuracy and by the time of the reading.
At ten fifteen, a student notices that the display has shown exactly the same value through a power interruption. A maintenance log confirms that the interface froze at nine fifty. This is an undercutting defeater: the present display is no longer a reliable report of the present temperature. It does not establish that the actual temperature is above or below twenty. The appropriate immediate revision is to suspend confidence in the current reading and take a fresh measurement.
A separate functioning instrument then reads twenty-seven degrees at ten twenty. Under the case's stated calibration conditions, that result supports a current temperature inconsistent with the earlier twenty-degree estimate. The times still matter. It does not prove that the greenhouse was already at twenty-seven at ten. A good incident note distinguishes the old reading, the discovered freeze, and the later independent measurement instead of merging them into one timeless fact.
The case shows how justification can change while the object under discussion remains the same. A previously reasonable route loses support when a specific defect is discovered. Investigators do not need to decide that every instrument is unreliable, or that the technician was irrational from the beginning. They need to identify the defect, replace the damaged evidential route, and express the new conclusion at the precision the replacement evidence permits. This is the practical value of distinguishing rebutting from undercutting information.
A defective instrument does not prove the opposite of its reading. If a scale is broken, a displayed five kilograms may be unsupported; the object might still weigh five kilograms. The defect undercuts the measurement. A separate accurate measurement of eight kilograms supplies a rebuttal under the stated tolerance. Keep these evidential jobs distinct, and do not turn a reasonable earlier belief into a claim that error was impossible.
State the claim under review.
p: the rod is ten centimeters long within the stated tolerance
The tolerance limits the precision being claimed.
Identify the offered evidence.
A checked ruler gives a ten-centimeter reading
A measurement is additional information, not repetition of p.
State the connecting assumption.
The ruler is aligned and its markings are accurate
These conditions connect the observed mark to the rod's length.
Check the supplied case details.
Alignment and calibration are stipulated
No known failure condition appears in this bounded case.
Record the supported conclusion.
The reading supports p within tolerance
Support does not license extra decimal places or an unlimited guarantee.
Separate the two propositions.
The card looks red; the card is red in daylight
A report of appearance has a different target from stable object color.
Record the observation.
The card looks red under the lamp
The experience supplies evidence only through assumptions about illumination.
Introduce the known condition.
The lamp emits red light
This can make non-red objects appear red.
Classify the new information.
undercutting
The lighting weakens the route without proving the card is not red.
Specify an appropriate check.
Inspect the card in standard daylight
A changed viewing condition directly tests the disputed connection.
Fix the original proposition.
p: the storeroom was empty at noon
The time is part of what must be supported or rebutted.
Name the original reason.
An assistant recalls an empty room
Memory is the actual support offered for p.
Examine the additional record.
A verified noon photograph shows a crate
The image concerns the same room and time as p.
State the photograph's implication.
The room was not empty at noon
One crate is enough to contradict the emptiness claim.
Classify the defeater.
rebutting
The new evidence supports the negation, not merely doubt about memory.
Bound the correction.
Revise the noon emptiness claim
The evidence does not show that all of the assistant's memories are wrong.
The claim p says this jar contains sugar.
A label is the offered reason
The label must belong to this jar to support its contents claim.
The label is known to have been moved randomly.
undercutting
This breaks the identification link without establishing different contents.
State whether not-p follows from the label defect.
p is supported only by a meter. New information says the meter is disconnected, without giving the actual value. Write R or U, then Y or N. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
| Response | |
|---|---|
| Kind | |
| Opposite |
p says no seeds sprouted. An authenticated photograph from the same tray and time shows a sprout. Complete the analysis. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Evidence symbols: R = rebutting; U = undercutting. Negation notation: ~P means not P; (P -> Q) means if P then Q. Negating a whole conditional is ~(P -> Q). Substitute the proposition letters supplied in the case.
Compare the new evidence with p.
The evidence supports opposite
An existing sprout contradicts the claim that none sprouted.
Name the evidential role.
The defeater is kind
The information supplies a contrary result rather than merely questioning the camera.
Keep the scope of the correction.
Only the no-sprouts claim is directly defeated
The photograph does not settle how many other seeds will sprout later.
p says a box is empty. A verified inspection finds a stone inside that same box at the relevant time. Write R or U, then Y or N. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
The defeater is kind. Does it by itself support not-p? opposite.
p is supported by a remembered date. New information establishes only that the memory came from a different event. Write R or U, then Y or N. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
The defeater is kind. Does it by itself support not-p? opposite.
p says a box is empty. Evidence A establishes only that its photograph was mislabeled. Evidence B is a verified same-time inspection finding a key inside. Classify each as undercutting or rebutting, then state Y/N: does A alone establish not-p? Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
| Your reconstruction | |
|---|---|
| A | |
| B | |
| A establishes not-p |
p says a hall was unoccupied at noon. A verified entry-and-exit log establishes that two workers remained inside continuously from eleven to one. Write R or U, then Y or N. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
The defeater is kind. Does it by itself support not-p? opposite.
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
p is based on a photograph. It is discovered that the photograph depicts another location; nothing is learned about the target location. Write R or U, then Y or N. Use the following symbols in your response fields; these codes replace word answers and remain the same in every language. Response symbols: Y = yes; N = no. Evidence symbols: R = rebutting; U = undercutting.
The defeater is kind. Does it by itself support not-p? opposite.
Without rereading, explain how to identify a reason offered to support a belief. Give a fresh case and identify what would change your analysis.
10. A damaged label, step 3
no
A randomly placed label could accidentally name the contents correctly.