I'm also surprised by this. Not too issues in this need peer reviewing.... They are just there. On the website. In the API.

Anyway, they probably just want to buy some time before giving a full statement... I guess

Luckily it's getting some press coverage

retractionwatch.com/2025/11/11/b...

Delighted to see the work covered here!! Great write-up in my opinion.

👍 a pleasure

I truly wonder how many people noticed over the past 14 years that something is off :D

I think many people might have noticed that the "This article is cited by" section is sometimes ... a bit off. But not sure how many api users there are at all.

Thanks for pointing this out!

Yes, it seems that there are mismatches in these things even internally within Springer Nature.

Great to see Zotero community actively finding and patching these things where they can. :)

Thx for the read. Share if you care. Read the article, especially plots and screenshots.

scirate.com/arxiv/2511.0...

🙏

And most importantly, double check your References!!

📚Journals affected? Thousands

📝Articles affected? Hundreds of thousands or even millions

🧑‍🔬Authors affected? Millions

⏰Since when? Most probably since 2011, when the API was released.

This leads all studied scientific reference databases to have incorrect references...

including the journal's own websites!! (check "This article is cited by" section for any article 1)

Article 1s have multiple std devs more citations than others, across many years and sources.

So:

❌ No article number returned by API.

❗️"Number" 1 returned for almost all articles.

‼️ Page Number 1 returned for ALL articles.

✅A recipe for disaster? Yes

📜Wait but there is PAGE NUMBER, isn't that good?

Nope.

In online-only Springer Nature journals, ALL articles are assigned a starting page number of 1.

Also, many are assigned an Issue Number of 1 (see "number" field). Almost all these journals have only one issue per volume

What's going on????

Well, it turns out if you request article metadata in json format via the SprinerLink API of Springer Nature, then you get
- DOI
- title
- authors
- year
- volume
- page number
- ....

🤯 BUT NO ARTICLE NUMBER!!! 🤯

Another weird thing: if we plot the citation count histograms of Nature Communications articles published on January 2nd 2025, we see the following.

The one article having about 10x the citations of other articles of the same age?

Article #1.

I'm glad to say this recently found a home in PRL after a very enlightening publishing process :)

doi.org/10.1103/nwzw...

Of course I lost my poster along the way, so I improvised a bit :D

Preprint out at
arxiv.org/abs/2503.24213
Check it out :)

This helps to close the detection loophole in experiments, where up to here people had to do global post-selection to keep only rounds when all photons arrived.

Using heralding information from an SPDC source for example, we show how this detection loophole can be closed.