“There is no scientific consensus on climate. The proof is a letter signed by scientists who don’t believe in global warming.” This argument, in various forms, has come up periodically for the past 20 years. Again this week, this kind of document was sent to António Guterres, Secretary General of the United Nations.

In reality, this argument reflects confusion: a scientist who expresses an opinion on a subject is not the same as a scientist who publishes a study on that same subject.

As The Rumour Detector undoubtedly should remind us more in its fact checking, the construction of scientific knowledge is not based on the opinion of one or more scientists, but on the accumulation of studies on a subject. A consensus is established when several studies confirm the same conclusion in a specific field.

Whenever you hear the well-known number “97% of scientists agree” about climate, this refers to 97% of the scientists who have published a study on global warming. This refers to a general agreement, even if there is no unanimity.

The origin of this percentage comes from research by the Harvard University science historian Naomi Oreskes in 2004: out of 928 studies published by climatologists and containing the key words “global climate change”, none disagreed with the consensus on the human responsibility for global warming. Since 2004, at least 14 similar compilations have arrived at totals ranging between 91% and 100%, all depending on the level of expertise retained for inclusion in the compilation.

Therefore, not only is there consensus on human responsibility, but this consensus increased between 1991 and 2011, according to one of these compilations.

Under the same definition, it is customary to say that biologists have a consensus on evolution, or that oncologists have a consensus on the carcinogenic nature of tobacco. This does not prevent people who hold a science degree from continuing to deny these facts based on their personal opinions.