Does cyclopentadienyl iron dicarbonyl dimer have a metal-metal bond? Who's asking?

In a 2012 essay, Green, Green and Parkin noted that certain bridged carbonyl complexes, including [CpFe(CO)(2)](2) (Cp = eta(5)-cyclopentadienyl), are commonly depicted with an explicitly-drawn metal-metal bond in textbooks, and presented a rationale for why that constitutes a misrepresentation. Here I briefly summarize the history of that compound, in which Cotton played a major part, and try to show that there are at least equally valid justifications for keeping the M-M bond in the picture. More generally, I will argue against the idea of a single best'' representation of a complex molecule, by drawing upon analogies to literary translation. (C) 2014 Elsevier B.V. All rights reserved.