Trace a Claim to its Original Source
In the last chapter you learned to place any source on the provenance axis: primary if it is the original first-hand record, secondary if it points back to one. Now you’ll put that axis to work. Most claims you meet in real life do not arrive straight from the source. They reach you after passing through a headline, a social post, maybe a press release, each one a step away from where the claim actually started. This chapter teaches you to walk that trail backwards and find the original, because that original is what you’ll judge and cite.
Tell the Original Source From the Messenger
Picture this. You see a headline: “New study says a 20-minute daily walk cuts heart-disease risk.” Three different things are tangled together in that one line. There is the claim (walking cuts heart-disease risk), there is the messenger (the news outlet that wrote the headline), and somewhere behind both there is the original source (the actual study the headline is talking about). They are easy to confuse, but they are not the same, and only one of them is the evidence.
The original source is the first-hand record where the claim came from: the study that ran the experiment, the dataset that holds the numbers, the report that made the finding. Everything that comes after it, the news article, the summary, the social post, is a messenger. A messenger can be accurate, but it is still relaying someone else’s work, not producing the evidence itself. That is exactly the primary-versus-secondary split from the last chapter, now seen as a chain: the primary source sits at the start, and every secondary source is a later link pointing back toward it.
Why does the distinction matter so much? Because claims drift as they travel. Each messenger trims, simplifies, or sharpens the wording to fit a headline or a feed. A study that found “a modest reduction in one measured group over eight weeks” can arrive at you as “walking slashes heart disease.” When you judge or cite only the messenger, you inherit that drift without knowing it. Reach the original and you can check what was actually found, in what conditions, and how strongly it was stated.
Do not treat a news report of a study as if it were the study. Writing “a study proved X” when all you read was a headline about that study is one of the most common beginner mistakes. You are citing the messenger and calling it the evidence. Always click through to the original before you trust or quote the finding.
Trace a Claim Back to Where It Started
Let’s walk one claim all the way home so you can see the moves. Tracing is just following the “where did this come from?” question one link at a time until you cannot go any further back.
The claim as you meet it. A friend shares a post: “Dark chocolate lowers your blood pressure, it’s proven!” That is the claim, with no source attached. Step one is simply to notice you are at the end of a chain, not the start.
Find the messenger it came from. The post links to a health-news article: “Daily dark chocolate may lower blood pressure, study finds.” Useful, but notice the giveaway words, “study finds.” The article is telling you about research; it did not do the research. This is a secondary link.
Follow the article’s own source. Good articles link or name where they got the claim. This one links to a press release from a university, the short announcement a research institution publishes to publicize new work. A press release is still a messenger, usually written to make the finding sound exciting, so keep going.
Reach the original. The press release names the journal article: the actual study that ran the experiment and reported the results. You cannot trace it back any further, this is the first-hand record. That study is the original source, the primary link, and the only one worth citing as the evidence.
Now compare the ends of the chain. The friend’s post said “proven.” The study itself might say “a small average drop in one group over a few weeks, with limits noted.” Same claim, very different strength, and you only see the honest version once you reach the original.
What if an article doesn’t link to its source at all?
Treat that as a warning sign, not a dead end. Search for a distinctive phrase from the claim, the named researcher, or the institution, and see whether an original study or dataset turns up. If nothing original exists anywhere, the claim may have no real source behind it, which is itself a finding worth recording.
Your Turn!
Time to trace a chain yourself.
The situation. A neighbour tells you: “Our town has the worst air quality in the whole country, it was in the news.” Curious, you look into it and reconstruct how the claim reached you. Here are the four links you find, listed in the order you discovered them:
Your task.
Identify which single link is the original source (the primary one you would cite).
Label each of the other three links as a messenger, and say in a few words what each one is doing.
Name one point in the chain where the claim could have drifted or been sharpened, and explain why.
Then check your answer against the model solution found at the end of this chapter.
Let’s Recap!
A claim usually reaches you through messengers; the original source is the first-hand record where the claim actually started.
Tracing means following “where did this come from?” one link at a time until you cannot go back any further.
Claims drift as they travel, so the original almost always states the finding more carefully than the headline does.
Citing a news report of a study as if it were the study is a common mistake; always reach the original before you trust or quote it.
News is usually secondary, but reporting based on a source’s own first-hand observation can be primary, so judge each link by looking.
You can now trace a claim back to the source worth citing. In the next chapter, you’ll see that finding the original is only half the job: you’ll learn to separate whether a source is relevant to your question from whether it is credible, and to keep a simple log of what you find.
Model Solution
Here is the annotated chain, walked from where the claim reached you back to its origin.
Link 1, your neighbour’s message. Messenger. This is the claim with no source attached, the very end of the chain. It tells you a claim exists, nothing more.
Link 2, the local news article. Messenger (secondary). The phrase “report says” is the giveaway: the outlet is relaying a report it did not produce. Useful mainly because it points you one step further back.
Link 3, the campaign group’s blog post. Messenger (secondary), and an interested one. It summarizes “a new report” and openly wants you to share its cause, so its wording may be sharpened. Still not the source.
Link 4, the environment agency’s dataset and report.Original source (primary). This is the first-hand record the others all point back to, and the one you would cite. You cannot trace the claim any further back than this.
Where the claim could have drifted: between Link 4 and Link 3. The agency’s report likely ranks the town on one specific measure over one period, with caveats. The campaign post compresses that into a flat “worst air, share this,” dropping the conditions. By the time it reached your neighbour as “worst in the country,” the nuance was gone.
And the nuance check from earlier: notice the news article (Link 2) here is secondary, because it relays the agency’s report. If that same outlet had sent reporters to measure the air themselves, that reporting would have been primary. You decide link by link, by looking, not by assuming “news is always secondary.”
