The Mellon Centre event looks interesting... but it's nowhere to be seen on their website (even though they have stuff on events happening much later, in July). Do you think it's because they are themselves deeply ambivalent about connoisseurship?

"[the basic skill of art history is] the ability to assign a date, place, and, if possible, a name on the evidence of style. I know no art historian who is not aware of the fact that this skill could not be practised in splendid isolation. The historian of art must be a historian, for without the ability to assess the historical evidence, inscriptions, documents, chronicles, and other primary sources the geographical and chronological distribution of styles could never have been mapped out in the first place."

I think there is a struggle to hand as the general run of art and art-history theorists believe that connoisseurship needs to be locked away in a cupboard (probably dark brown 18th century gothic revival) and not mentioned.

It’s an enormously important area and I wonder about finding another name for it so that ordinary folk don’t get frightened off... ?

In your recent blog you asked for views on connoisseurship. Perhaps an obvious point but one which does not seem to be stressed much is that science i.e. proof of facts such as pigment identification, dendrology  etc.  can only really be used to prove unequivocally that a painting is not by a given artist. (I do not include fingerprints or handwriting in this which are subjective fields and, speaking as a  lawyer,I know how woeful the track record for these is.  ). Whilst science  may contribute towards a positive identification it is hard to see how it could ever do so unequivocally on its own. So long as positive identification is desired therefore connoisseurship will be essential. Or am I being simple minded?!

Their isn't any certain recipe for attributing a work of art about which a doubt exists or should exist. Connoisseurship is a tool in authentication and attribution.  It isn't the only tool, but it is an essential one.

If one thinks of a hierarchy of authentication: first is a signed work with documents that show that it was by a particular artist, science that validates the materials, and a provenance that can trace this particular piece to the artist, which only leaves the possibility of intentional fraud by the owner who could have substituted a copy with the right materials for the original.

After that all of the tools of authentication must be applied to the work.

Provenance - documentation and historical support.

Scientific examination - of the pigments, canvas, wood, and other materials.

And then Connoisseurship.

In general, scientific data can only disprove an attribution.  It can only show that a work could not have been created by a particular artist or in an positive sense, that it might have been created by a particular artist.  Even a work on a piece of canvas cut from the same larger piece of canvas as a work by Vermeer could be (admittedly unlikely but still possible) by a contemporary, but for an artistic examination of the work itself.

If documentation is lacking and the work passes the other tests, and probably some which I have overlooked, connoisseurship is still necessary.

Two identical or similar works of the same vintage are often by two different artists and could pass other tests including provenance, both possibly having had the same original owner who wanted a copy, and ultimately it is style, brushwork, peculiarities of signature or other indicia (Strong noted how dates were indicated), and the other elements of connoisseurship that can attribute (provide an informed opinion regarding) the authorship.

Like science, connoisseruship isn't a proof of anything only a statement that a particular artist could have created, might have created, or is very likely to have created a work.  It can also suggest that there is evidence to disprove an attribution.

Then, when there is a individual work, aside from deliberate fraud which is a separate topic, there is the question of whose hand created it or which parts of it which, in the absence of other proof, requires connoisseurship.

But this still only an informed opinion.  A great difficulty with connoisseurship is assessing the agenda and qualifications of the expert.  There are both professional and financial pressures at work here.  And the expert is only a human.

Why consider connoisseurship, because the scientific  tools are also inconclusive, and C adds evidence to build an opinion.  The work, in general, must speak for itself.  Res ipsa loquitor.

As a painter the bit I'm always troubled by is something that is rarely spoken about in the art world - some experts are colour blind and some others have no spacial awareness and that's why for many it's easier to talk about scientific analysis and provenance, without looking at the picture concerned and asking basic questions about why, what's achieved and how.  

Sometimes with a collector one can tell what their strengths and weaknesses are from the art they collect and that's true also with gallery owners who choose and put on exhibitions - I can think of several who I think might be colour blind! (I won't name anyone)  But 'experts' are opaque about their abilities / prejudices and preferences.  And frequently they are unable to either explain decisions or engage in debate about these decisions.  I can understand why they back off - especially if the other party has a financial incentive to prove a picture.      

I'd happily try to devise an exam in practical skills in which experts could demonstrate their understanding and sensitivity to line, colour, texture, composition, sculptural qualities and pattern.  Maybe a if an expert could analysis - for example which colours he/she can see in a shade of grey or do a quick sketch to demonstrate their understanding of the work they are looking at, this would inspire greater confidence in them.