A blog devoted to exploring wines made from unusual grape varieties and/or grown in unfamiliar regions all over the world. All wines are purchased by me from shops in the Boston metro area or directly from wineries that I have visited. If a reviewed bottle is a free sample, that fact is acknowledged prior to the bottle's review. I do not receive any compensation from any of the wineries, wine shops or companies that I mention on the blog.

Thursday, November 29, 2012

Book Review - Wine Grapes, by Robinson, Harding & Vouillamoz

When I first learned about the impending publication of the massive Wine Grapes, I was both excited and a little nervous.  I was excited because it looked to be just the kind of thing that could help me a lot in doing research for this blog, but at the same time, I worried that its publication would make me obsolete.  The book purported to be the ultimate guide to wine grapes, and I worried that everything interesting that could be said about many of the grapes that I try would already be said here and there would be no reason for me to continue with this site.  I pre-ordered the book many months in advance and began the waiting process.  That process ended a few months ago with the release (and delivery) of this massive tome, which I've now had some time to read through and put some thoughts together about.

Those who are unfamiliar with some of my other posts should note that I've had some issues in the past with the crew responsible for publishing this book.  This is the same editorial team that is responsible for maintaining the online version of the Oxford Companion to Wine, and many of the findings from this book were previously published through that venue.  In several of my prior posts, I've noted issues or problems that I've found in the online version of the OCW, some of which I addressed to the editors, but none of which were really satisfactorily resolved.  The first involved their entry on Hondarrabi Zuri (link goes to my discussion of the issue), where they claimed that it was actually the same as the American hybrid Noah, while the second involved their publication of an alternate parentage for Emerald Riesling, which turned out to be correct, but which they refused to divulge their source for.  I'm not going to delve back into either of those topics, as I've said all I card to say in those posts, so interested readers are advised to follow the links and read away.

Before I get into the issues I have with this new book, I do want to start out with a list of the things that I really liked about it.

1) It is gorgeous.  The layout is easy to follow and though there is a lot of information presented for most of the grapes, the book never feels overwhelming.  The slipcase is a nice touch as well and I much prefer the burgundy color for the US publication to the cream color for the UK.

2) Most of what is presented is accurate.  The book is thoroughly researched not only from a scientific standpoint, but from a historical and ampelographical one as well.  Much of the information provided is adequately cited and the bibliography is comprehensive.

3) The scope of the work is astounding.  It claims to be a "complete guide to 1,368 vine varieties," and while I didn't count to make sure they're telling the truth, the 7 pound weight and nearly 1,250 pages would seem to indicate that they are.  There is no other single volume available with more information on more grape varieties than this.  No really serious wine adventurer will want to be without this volume for very long.

4) The signed bookplates that the group gave away was a very nice touch.  All you had to do was email your address to someone affiliated with the publication team and they mailed out a book plate signed by all three authors for you to affix to the inside cover of your tome (for free!).  If you haven't ordered the book yet or haven't yet asked for a signed bookplate, you're probably too late, I'm afraid to say, as they only did a few hundred of them.  I'm happy to have received mine and think it was a nice way for the contributors to thank people for laying out some serious cash for this book.

All that said, there are some issues I have with the book.

1) The color plates.  There are 5 sections within the book that contain 8-10 pages of full color reprints of illustrations from a turn of the century (last century) work of ampelography.  These sections are evenly spaced throughout the book.  They're printed on a different kind of paper that is actually a different size (smaller) than the rest of the pages, and when you try to flip through the book, you inevitably land on one of these sections.  They're nice to look at, but they're completely unnecessary and kind of in the way.  The book retails for $175, and one wonders how much of that they could have shaved off by doing away with these.

2) It is sometimes difficult to find exactly the grape you're looking for.  They put a grape's main entry under its most common synonym in the country that is thought to be its ultimate home.  This is fine for most grapes, but you end up with some weird things like having Carignan listed under Mazuelo, which is a synonym I'm only familiar with in Rioja.  The index is helpful for navigating around this issue (which was probably inevitable given the book's scope), but flipping back and forth in a 1,250 page work is no picnic.  There are some guidepost entries throughout the text that redirect you to the proper name, but these are somewhat erratically placed.

3) The research.  This is my biggest issue, and it's what I'm going to spend the bulk of the rest of this review talking about.  If you don't care about it, feel free to close this browser window now and you'll be welcome back when I get back to reviewing unusual wines in the next few days.

The problems I have with the research for this volume fall into two broad categories.  The first category concerns research done by others that is reported in Wine Grapes.  As I mentioned earlier, most of the information in the book is accurate and is well researched, but there are a few places where the research is either sloppy or inaccurate for personal reasons.

For example, I recently wrote about the Fetească Albă grape from Romania, and in that post I lamented that the OCW mentioned that the long-accepted synonymy between this grape and the Leányka of Hungary was inaccurate, but I had some difficulty tracking down their source for the information since it was only cryptically mentioned and not cited.  This assertion is reprinted in Wine Grapes, and 2 citations are given.  One is the paper that I mentioned in my post on Fetească Albă, which has some serious methodological issues and can hardly be taken to be indisputable (please see that post for more details).  The other is this paper from a Hungarian team, which, while interesting, does not actually say anything at all about the relationship between Fetească Albă and Leányka.  It does say that Leányka and a grape called Leányszolo are not identical, but there is no mention at all of Fetească Albă anywhere in the paper.  Citations are great, but only if the research that you're pointing to actually says what you're claiming it says.

As for the charge of inaccuracies due to personal reasons, the example of Ciliegiolo and Sangiovese provides a nice example.  As mentioned in my previous post on the Ciliegiolo grape, the exact relationship between Ciliegiolo and Sangiovese is contentious.  It is known that the two grapes have a parent/offspring relationship, but there are two different theories as to which is the parent and which is the offspring.  The first was reported by Jose Vouillamoz (and several other Italian scientists) in 2007 and they claim that Sangiovese's parents are Ciliegiolo and a grape called Calabrese di Montenuovo (which Vouillamoz & crew supposedly just stumbled across somewhere in Calabria...it is unclear whether it exists in any holding facilities or even whether any other research teams have been able to obtain a sample to test for themselves, all of which will lead into my final issue in the following paragraph).  The second (authored by Di Vecchi Staraz and others, full citations are in my Ciliegiolo post linked above) shows that Sangiovese and a grape called Muscat Rouge de Madére are the parents of Ciliegiolo.  Vouillamoz's results have not been duplicated by any other group, but Di Vecchi Staraz's findings have been duplicated in at least 2 different studies.  Wine Grapes, whose scientific content was edited by Vouillmoz, presents Vouillamoz's findings as if they were the final word on the matter.  The book does make mention of Di Vecchi Staraz's findings, but dismisses them because they "do not fit with the parentage of Sangiovese (Vouillamoz, Monaco et al 2007)," which is question begging of the worst kind.  The book does note that there were other "discrepancies" in Di Vecchi Staraz's findings, but I've not seen those discrepancies indicated anywhere else, and, as mentioned above, there have been 2 studies in the past year that have shown that Di Vecchi Staraz's analysis is correct.*

All of which brings me to my major second concern with some of the research presented in Wine Grapes, which is that it is private research.  There are many posts where the citation for the information provided is given simply as (Vouillamoz).  Vouillamoz has several publications listed in the bibliography of the book, and when one of them is referenced, the year is always given.  These (Vouillamoz) citations represent something different, though, which isn't ever really fully explained in the text.  Vouillamoz maintains a private DNA microsatellite database with information on various grapes that he has either analyzed in the course of his research, or which may have come from other studies that he has read.  Much of the stuff that the OCW has published, and many of the relationships published in Wine Grapes, were "discovered" by Vouillamoz by sifting through the information in his private database, or by Vouillamoz conducting experiments on his own and publishing the results in Wine Grapes for the first time.

As an example, Wine Grapes publishes that Listán Prieto, aka Mission, and Listán Negro from the Canary Islands are actually two different grapes, despite the fact that they have long been considered to be the same variety. Vouillamoz claims that a "study" has shown that Listán Negro actually has a distinct DNA profile, but his citation is "(Vouillamoz; Jorge Zerolo, personal communication)," neither of which can be called a "study" in any meaningful sense.  The information is presented, though, in a way that makes it seem like the finding has been verified by DNA research (there is an icon in the book, that is probably supposed to look like a double-helix but which just looks like a circle with open semicircles on its top and bottom, used to represent when a synonym or a homonym has been "verified" by DNA studies, and that icon is used here), when in fact it seems that it has only been seen in one non-published "experiment."

Here's the problem with this: it isn't science**.  The information that Vouillamoz chooses to publish in this way may be interesting, but it's wrong to consider these findings "scientific" in any way.  There's a process that people have to go through in order for their ideas to be accepted by the wider scientific community.  That process involves not only performing the experiments and analyzing the results, but publishing them in a peer-reviewed journal as well.  If you read through any of the papers I link to on this site, you'll see that they aren't just a guy asking a question and then announcing that he's answered the question.  The paper goes through the entire set-up for the experiment and shows the data obtained from the experiment described.  This process exists so that other scientists (or anybody, really) can analyze your methodology and your results to see whether your experiment actually can give the kind of information that you claim it can and whether it actually does give the information that you claim it does.  The foundation of good science is reproducibility, which means that others should be able to read your paper and perform your experiments themselves to see whether they obtain the same results as you.  Just publishing findings doesn't fly.  You're not right when you make a discovery; you're right when someone else verifies your results.***

I'm not suggesting Vouillamoz doesn't know what he's doing or that he has fabricated results.  What I am saying is that any information that he provides from private research or from any source that is not a scientific journal is unverified.  My issue isn't the publication of unverified findings, but rather that this information is presented to the reader in such a way that is difficult to separate from the more thoroughly researched findings.  It is given to the reader in a format that very closely resembles properly cited findings, so the lazy reader notes the parentheses and moves on, seeing a citation where none exists.  When citing the published results of others, readers can consult that source and decide for themselves whether the findings seem accurate given the description of the experiment and the publication of the data.  This empty citation, though, leads nowhere and does not even have a corresponding entry in the Bibliography.  It is telling us that Vouillamoz did something, but there's no way for us to see what it was.  By just giving the conclusion for this private research, the book is essentially asking us to take Vouillamoz's word that he did everything correctly and that his experiments and/or his interpretations are sound.  Given some of the issues I've encountered previously in his work, that doesn't feel like a safe assumption to make.

I'm not saying that Wine Grapes is worthless because of these issues.  As I said before, the overwhelming majority of the information in the book is accurate, interesting and informative.  What I am saying is that some of the information they present as if it were scientific fact is not that at all and should be taken with a grain of salt.  I feel that the book could have been better if some of the controversies were more thoroughly explained or if some of the posts had less of the authoritarian air that they carry.  Jose Vouillamoz is a good scientist, but asking a single scientist to compile all of the information for this book was probably not the best idea.  What you get is a single expert's bias, which doesn't produce a very scientific result.

My bottom line is that this book is now my first line of research for pretty much every post I'll write from here on out.  It's an incredible resource that I look forward to enjoying for many years.  But like any first line resource, I'm not going to be content to stop my research there, and I will pursue any questions I have as far as I can, or until I feel that they have been thoroughly answered.  Grape genetics is a rapidly moving field that I enjoy following along on the sidelines.  I think in about another decade, the 2nd or 3rd edition of this work may end up being definitive, but we're just not quite there yet.

*The first is mentioned in my post on Ciliegiolo. Citation is: Cipriani, G. et al. The SSR-based molecular profile of 1005 grapevine (Vitis vinifera L.) accessions uncovers new synonymy and parentages, and reveals a large admixture amongst varieties of different geographic origin. 2010. Theoretical and Applied Genetics. 121: 1569-1585.

The second is Lacombe, T., Boursiquot, J.M., Laucou, V., Di Vecchi-Staraz, M., Peros, J.P., & This, P. 2012. Large scale parentage analysis in an extended set of grapevine cultivars (Vitis vinifera L.). Theoretical and Applied Genetics. In press.

**You may argue that I'm missing the point here.  This book is for entertainment purposes and is not intended to be a scientific treatise, you may say.  And you may be right.  But the book is being marketed in such a way and its content is arranged in such a way that the authors clearly want you to think that what they are publishing has been "scientifically proven."  I am arguing that this insinuation is dishonest on their part, as much of the data they are publishing is unproven in a scientific sense. Think I'm overstating their case? The following is the publisher's blurb for Wine Grapes from Amazon.com:

"Using cutting-edge DNA analysis and detailing almost 1,400 distinct grape varieties, as well as myriad correct (and incorrect) synonyms, this book examines grapes and wine as never before. Here is a complete, alphabetically presented profile of all grape varieties of relevance to the wine lover, charting the relationships between them and including unique and astounding family trees, their characteristics in the vineyard, and—most important—what the wines made from them taste like."

Their claim is that they have used "science" or some "science-like substance" to show the truth about many grapes and how they are or are not related to one another.  All I am trying to show is that their methodology isn't always sound, and as a result, some of the conclusions that they reach are hardly as definitive as the confident tone of the blurb and the text itself would lead you to believe.

***As I've pointed out numerous times on this site and a few times in this post, this isn't insignificant stuff (well, it may be in a cosmic sense, but you know what I mean).  Vouillamoz's claim regarding Fetească Albă and Leányka comes from a paper that used samples from a holding facility known to have labeling errors.  This is a methodological issue that requires follow-up research, but we wouldn't know that without the full published paper.  I don't believe that the follow-up research has been done, which only means that the question is currently open until proven otherwise (but if forced to make a decision, you're probably wiser staying away from the findings of a flawed study and opting instead for the belief that the prior hypothesis was correct).  Additionally, Vouillamoz's claim about the relationship between Ciliegiolo and Sangiovese has not been replicated by any other group, and it is unclear whether it can be because of the special nature of one of the purported parents.  If other groups cannot test Calabrese di Montenuovo themselves (and I do not know whether they can or not, but no one else has, as far as I can tell), then we just have to take Vouillamoz's word for it that the grape that he happened to stumble across (he himself says it was "fortuitously sampled") is the long-lost parent of one of the most famous wine grapes in the world. The thing is, we don't have to take his word for it, because we have information from three other publications from three other labs that have all independently shown that Vouillamoz's findings cannot possibly be right, since Sangiovese is itself a parent to Ciliegiolo and is not one of its offspring.  Their methodologies and results relied on grapes and samples that other labs could access and test themselves rather than on one-off field samples and private databases.

UPDATE: I uncovered a bit more on the Ciliegiolo/Sangiovese issue recently, and posted my findings in the comments to this post.  You can read the relevant information here.  

9 comments:

Andrew said...

Really good review.

I enjoy the book a lot and do disagree on the pictures. They add something for me, thought I can see how it might not be worth the extra cost.

The main reason I comment is that I wholly concur with the rest of your assessment. I had been mulling those same thoughts, but had not been able to crystalize them. Thanks.

It does seem that science has taken a back seat in this field. I suspect that it is due to there been significant money (and prestige) to be had by holding access to knowledge and resources vs 'open source' methodology. Alas, that openess is required for science and it is being sacrificed to commercial ends. I don't really fault any of the parties and I agree that none of them appear to be dishonest or ill-intentioned.

The last comment I would make is that while this immense tome is cool in a way, the knowledge in it and the interchange is really best served by online resources which can be updated with new information, retracted information and directly linked to the sources.

Kevin said...

Thanks for taking the time to read, analyze, and review this tome. While I doubt I'll ever pick up a copy myself, I feel quite enriched for you having done so and being willing to go beyond the limitations you cite in writing your blog entries. Cheers!

Bibulous said...

Bravo for standing up for scientific rigor! In the ideal future there will be a public online database containing all available grape DNA analyses (and a guide to understanding them). Until then, Fringe Wine will continue to be a trustworthy starting point for amateur variety hunters. I would like to mention one aspect of "Wine Grapes" that makes it invaluable for us explorers. Nearly every entry has a "where it's grown and what the grape tastes like" section, and very many of these sections name producers of varietal wines made from that grape. That plus Google empowers anyone (with an unlimited budget and travel time) to taste these wines, which after all is our chief goal.

Andrew said...

One additional comment : The little color icons are the dumbest waste of ink. It took me some serious dedicated searching to find grapes that used just off-white or just off-dark red.

WineKnurd said...

As a scientist myself, it is of critical importance that all conclusions presented in the pretext of science be subject to proper and conscionable scientific review. As far as I can tell, this book has been published under the pretext of science and also edited by a scientist. It's very easy to see that Vouillamoz is using this volume as way around the scientific process; he presents "science" in an authoritative tome which gets propagated as other scientists cite "Wine Grapes", thinking that his citations have been properly reviewed by peers. This agenda is rather abhorrant and I have to disagree with Andrew's comment that this was anything but conscious dishonesty on the part of Vouillamoz. What Vouillamoz never expected was the "great effort" that Fringe Wine put into checking the citations (i.e., just reading them). Vouillamoz is promoting his agenda and looking to make a reputation in an entirely dishonest manner. It is also very clear that he is intended to be the scientific expert of the book, as Harding and Robinson have no scientific background. Who was going to double check the science, since it was his job to double check these facts? I am very glad that you called him out on FW, and hope that this post shows up on every google search which includes Vouillamoz or Wine Grapes.

Fringe Wine said...

Thanks everyone for your comments. I'm gratified by everyone's support, since I was unsure how this review would be taken given the glowing, fawning reviews that were popping up everywhere else.

@Bibulous: Your comment about the producers given for some of these grapes is a good one, and is something I neglected to mention because I live in a state in the US that prohibits direct shipping from out of state. I have been able to obtain some wines through various channels as a result of the information in this book, but it's not as valuable as it could be because of the antiquarian laws still on the books in an awful lot of places in the US. This is obviously not the book's fault, but it is an unfortunate consideration for too many of us.

@Andrew: The illustrations seem to be a contentious issue, as I've heard from several people who find them useful and several others who find them obnoxious and intrusive. I'd be less inclined to hate them if they were the same size as the other pages, but I find that they get in the way more often than they contribute. Also, if I'm not mistaken, the images themselves are public domain, so their acquisition cost the team nothing, but printing them probably contributed in a not insignificant way to the final cost of the book, which is borderline prohibitive as it is. Add to that the fact that they're about 100 years old, which causes one to wonder how accurate they may be, and I just would feel better if they weren't there. I agree completely about the colo(u)r scale, though. A simple white/pink/black indication would have been more than sufficient, and could have been provided textually (W/P/B, etc) rather than with an actual color scale.

@WineKnurd: As always, thanks for your engagement and your comments. I appreciate your input as a scientist and am gratified that you agree with many of my criticisms. Your comment helpfully expanded many of the issues I had that I was only able to hint at.

Thanks again to everyone for reading and commenting, and I look forward to hearing from anyone with opinions on the book itself or with my specific criticisms of it.

Emily H. said...

Thanks for your critical view of this (relatively) exhaustive publication. As always, we really appreciate the time and effort that you have put into this research so that other wine and vine enthusiasts can benefit. Did you find any changes to the entry for Hondarrabi Zuri, or was the same erroneous information still there? Keep up the good work and thank you for providing beneficial and constructive criticism in this emerging field of research :)

Rob Tebeau said...

Hi Emily:

There actually is no entry in the book for Hondarrabi Zuri. When you flip to where it should be alphabetically, there's a redirect entry telling you to look at the entries for Courbu Blanc, Crouchen and Noah. The most substantial treatment of the grape is given in the Courbu Blanc entry, which does indicate that Courbu Blanc = Hondarrabi Zuri. The entry mentions that the name Hondarrbi Zuri was also used for Crouchen, and that "another reference sample of Hondarrabi Zuri is in fact identical to Noah, an American hybrid."

All of which is technically correct, if not very clear. I still think that most of what is called Hondarrabi Zuri is actually Courbu Blanc, though some of it could be Crouchen. The issue with Noah is due to a mislabeling at a holding institution, and that is the "reference sample" referred to in the quote above. The entry on Noah offers a bit of extra information, but only in the context of explaining the findings of some of the papers who got their hands on the bad Hondarrabi Zuri samples.

Given that it's essentially an issue that is only of interest to those who are really closely following the literature, I'm not sure why they'd even bring it up. If they felt the need to do so in the interest of completeness, something like the following would have been sufficient: "Though some recent studies have shown that Hondarrabi Zuri is genetically identical to the American hybrid Noah, this finding is due to a mislabeled plant in a university holding facility, and it is extremely unlikely that any wines marketed or labeled as containing Hondarrabi Zuri are made from Noah grapes."

Rob Tebeau said...

I also wanted to post a follow up comment on the Ciliegiolo/Sangiovese issue. As mentioned above, Vouillamoz dismisses the findings of Di Vecchi Staraz et al because of "discrepancies" in their DNA data that would warrant further investigation. I didn't remember any discrepancies in the paper, but I pulled it back out again and re-read it to see if I had missed something.

What the paper says is that the team was able to analyze their parentage claims using 38 different sites. The only discrepancy that they found was that at one of the sites, one of the DNA copies of one of the grapes didn't quite fit with the proposed parentage (meaning that between the 3 grapes, 5 of the 6 sets of DNA were consistent (possibly identical, it's hard to parse the sentence), but one set was different). Discrepancies of a single base pair difference are generally ignored, as these are relatively common mutations, but more base pairs than that, and you've got to take a closer look.

It is usually the case that one discrepancy in a parentage analysis is enough to toss out the whole proposed pair, but usually if there is one discrepancy, there are other discrepancies, so in practice, it doesn't often come up that an entire pedigree is tossed due to a single mismatch. When it did come up in a study published in 2003, where there was a single discrepancy that was 10 base pairs long, the research team ran the numbers and decided that this was probably due to some kind of unusual mutation (such as chimerism), and that the parentage analysis was OK despite that one issue (exact quote: "a sole multiple repeat unit discrepancy is not sufficient to reject a parentage"). Their reasoning was sound and the numbers seemed to back them up, and other scientists have since followed the lead of this team. It is now relatively uncontroversial that a single "null allele" cannot scrap an entire pedigree proposal if the match is good at enough other sites.

Anybody have a guess as to who the lead author was on that paper in 2003?

If you guessed Jose Vouillamoz, you get a gold star.

So the "discrepancies" (sic, as there's just the one) that Vouillamoz is lamenting in the Wine Grapes book is actually justifiable using some of Vouillamoz's own prior research. The problem is that finding the parents of Sangiovese is a huge deal, and Vouillamoz doesn't seem to want to admit that his results showing Ciliegiolo as one of the parents has been disproved. It feels like his personal agenda is getting in the way of the truth, and as I mentioned above, that's really lousy science.

Vouillamoz, J, Maigre, D, & Meredith, CP. 2003. Microsatellite analysis of ancient alpine grape cultivars: pedigree reconstruction of Vitis vinifera L. 'Cornalin du Valais.' Theoretical and Applied Genetics. 107, pp 448-454.