Colon: A Plague

Google trains the writing of headlines, which exchange creativity for audience

Mario Sergio Conti, last weekend, wrote about the semicolon’s risk of extinction. He also recalled that "exclamation points abound, buried in the newspaper by the swashbuckling columnists." I add another pesky abundance to the analysis: the colon sign.

As well as the imperatives "understand" and "see," the colons are in many titles of Folha and other vehicles for brevity, in principle. Word or expression, which identifies the subject of the text, is separated by a colon from the information to be transmitted. An example in the last week was "Datafolha:", used to exhaustion. "Datafolha: Bolsonaro voter distrusts the polls, defends weapons and sees optimism in the economy"; "Datafolha: 7 out of 10 reject the idea that guns bring more security"; "Datafolha: Aversion to Bolsonaro is dominant among poor and rich women".

On the latter, a comment on the Reader's Panel even played with the format. "Women: always superior."

The statement "98% want the Covid vaccine to remain free to all, says Datafolha" is the exception of the last few days that explains the unwritten rule. Avoiding "says," "sees," "reveals" is to save precious touches to quote everything that the institute finds. Furthermore, in the flat world of the internet, where only headlines are read, the so-called hat, that keyword that comes above the utterance, is almost always ignored. A serious problem with newspapers is to make it clear that some articles are of opinion and humor. Then the solution is "Opinion: GDP has improved, but few people felt it in their pocket, skin or soul."

Things start to get complicated when the system contaminates headline where the marker is not needed. "Pelé uses Ukraine game to ask Putin: 'Stop this invasion'." This one arrived in print, on Thursday (2), in collision with the Editorial Manual (page 125). The last item on headlines asks to avoid exclamation (except in exceptional situations), period, colon, question mark, ellipsis, dash and parentheses. The newsroom, however, uses another instruction on the same page of the Manual, which guides, on digital platforms, the use of the keyword "positioned at the beginning of any headline."

All this would just be curiosity if the end result were not boredom. Due to the form of the titles, sometimes necessary, but mainly due to the blatant lack of creativity. And the criticism here is not for journalists. Search systems, ie Google, train the writing of statements, dictate the audience formulas. Between the genius title and the most searchable, anyone will choose the one that pays the bills.

To imagine that such influence of search engines is only in the title is naive.

DEPP V. HEARD

Johnny Depp sued his ex-wife for an article she published in the Washington Post in 2018 in which his name is not mentioned. After a planetary carnival, which many followed and many ignored, the jury questioned which was one of the objects of trial, everything she had written or just the title. The judge responded that it was to take the headline into account: "Amber Heard: I spoke out against sexual violence — and faced the fury of our culture. This has to change."

With a colon, of course.

AGGREGATES

Days after Folha brought the most important poll of the year so far, the one that points to the chance of an election decided in the first round, its direct competitor announced an aggregator of surveys. Estado de São Paulo uses data from 14 companies to predict "the most likely scenario of the dispute each day."

Aggregators are common in the US, but new to the general Brazilian public. The secret lies in the calibration of the algorithm that balances the different searches, assigning different weights according to methodology, field size and other characteristics. Financial market agents, which currently sponsor most of the institutes, use similar tools for private analyses.

The burden of seeking a weighted average of the polls is to miss the news. Datafolha was the news of the last week precisely because of the result it brought alone. Had its numbers been diluted with others, which did not capture the same movement of voters, it would not be on the lips of the political world, which moved a lot in the face of the perspective of the first round.

Nevertheless, there are merits in the conjoint analysis, as long as it is understood as a data study, not an effective research. Which, obviously, will not happen, as can be seen in the titles of Estadão and other vehicles with aggregators, such as the Jota website, where the tools gain the comfortable role of their own research. The expensive budget of others is a mere detail.

Translated by Kiratiana Freelon