Northerners used to stop and fill their 20-liter jugs on their way home from the Algarve. Now, they buy bag-in-box from their local supermarkets. The modern taste for fresher fruitiness and big oak hasn’t helped, either. The talhas in his annex sit empty now, with the tasca’s working pots making just enough to carry his customers through the next vintage. Tascas are a dying breed.

But still, every September, Antonio’s café closes down and the talhas are cleaned and resurfaced with a fresh mixture of pês (pronounced pesh) to prevent leakage (see also below). Soon the new batch of grapes arrives from the surrounding vineyards to be crushed, with both must and skins filling the talha. Fermentation burbles away naturally for a couple of weeks, and when finished, a micro-thin layer of olive oil is added to seal out oxygen. No need for sulfur dioxide or filtration, this is winemaking’s answer to “set and forget,” one-pot slow cooking.

Gato, and other villagers making wine only for family use, leave the wine to mature until St Martin’s Day, November 11. Then, the local priest leads a procession through town, ending with a blessing of the new wine. Soon after, talhas are tapped for the first taste of the vintage, and the ensuing festivities carry on into the early morning.

Ancient traditions

My first encounter with “amphora”-made wine happened years ago in Campania (note here the incorrect use of the term amphora, which is actually a long, slender clay pot with handles, used only for transportation, not fermentation). In Campania, Luigi Tecce was puzzling how to make Aglianico in terracotta, following both Ancient Greek and Roman recipes. His experimental Aglianico was as good as anything I’d had from conventionally made Vulture or Taurasi—more interesting, actually. 

But like many modern attempts at terracotta winemaking I’ve observed firsthand since then (in Istria, Croatia, Slovenia, Friuli, South Africa...), his work was reliant on vague instructions from ancient texts, guesswork, and a whole lot of trial and error. Antonio Gato’s wine, on the other hand, was like a living fossil, with clear links straight back to the earliest days of winemaking.

The Georgians, too, have an unbroken multi-thousand-year history of clay-pot winemaking in a vessel called a qvevri. But the northern Caucasus’s qvevri traditions differ fundamentally from the southern Mediterranean’s talha traditions. While their terracotta technology and winemaking practices share many similarities, there are significant differences as well, resulting in different wine styles. Buried underground, qvevri are sealed with damp clay and eventually drained with a bucket from the top. Talhas are free-standing and exposed to the atmosphere, sealed differently, and drained by gravity from taps below.

Unfortunately, while Georgian qvevri traditions remain very much alive and predominant within Georgian culture, Portugal’s ancient unbroken talha traditions are considerably less secure. When I first began searching out talha-made wines back in 2012, there were just a handful of active tascas and family producers, another handful of young winemakers focused on Alentejo’s relatively new Talha DOC, and one quixotic attempt to resurrect factory-scale production—all desperately trying to keep talha traditions alive. It was a precarious situation, with talha barely holding on by its proverbial fingernails.

Ground Zero

About an hour’s drive south of Arcos sit the ruins of Iberia’s largest Roman-era farming complex, Villa Romana de S Cucufate. Neolithic stone circles and unbarrowed, henge-like Dolmen (antas in Portuguese) dot the landscape, with many more rolled away and piled up centuries ago to make way for plowing. A few days’ walk south, where Iberia narrows within sight of North Africa, early waves of hunters and gatherers once paddled across the Mediterranean on their migration north. Later, the area was inhabited by the Tartessian culture, known for its wine. Iron Age jewelry, decorated with grape bunches, was excavated locally. The place is old.

S Cucufate straddles an ancient place that has been making wine for a long, long time. The Phoenicians and Greeks were there, and the Romans eventually made large quantities of wine in 1,000-liter talhas between the 1st and 4th centuries—the Gallo of its day? By the Middle Ages, Christian monks took over production and renamed the place for their favorite saint, Cucufate. Within short walking distance of S Cucufate, the people of Vila de Frades, Vidigueira, and other nearby villages cling tightly to their talha and tasca traditions.

A key reference point for talhas in Vila da Frades, if not the region, is Professor Arlindo Ruivo. He is a font of practical knowledge and oral history concerning local talha production stretching back deep into the 20th century. Much of this was gathered first-hand, making talha wine for his family’s business. Semi-retired now, he keeps his hand in, producing wine for family consumption and helping youngsters renew the tradition. 

A visit to his wonderfully cluttered, rust- and dust-filled cellar, packed with essential paraphernalia of talha production, is a joy to the eyes. Purpose-built in the 17th century as a producing cellar, it has vaulted ceilings that echo those in nearby S Cucufate. Deceptively larger than it seemed initially, I counted roughly 40 large talhas tucked away in its cool, dark corners, suggesting peak production was around 50,000 liters. Most of this would have been made from white Antão Vaz grapes and sold in small barrels to restaurants and shops throughout the Beja region.

My first question concerned the mysterious stuff known as pês—a lost art it seems, since no one makes it now or agrees how it was made. From my time in Georgia, I knew their version there was a simple coating of beeswax applied shortly after firing their qvevri. Because both qvevri and talhas use raw terracotta, there seems a conscious rejection of glazing’s more airtight seal in favor of the micro-ox transfer pês offers.

Traditional pês was clearly a complex concoction. Ruivo remembers the time when almost every family in the village had a talha with their own secret recipe. It was possible to taste a wine then, and from its pês character, guess who made it. His recollection is that the basis was rosemary, thyme, pine-nut tree resin, and olive oil, carefully boiled together to a point where it wasn’t too liquid or too solid. This Goldilocks Point ensured resin characters didn’t enter the wine.

Gato also remembered specialist pesgadores who moved from village to village inverting talhas over fires to melt out the old pês linings, then scrubbing the talhas with fennel and recoating. Complicated, but much easier than cleaning buried qvevri.

A new Roman villa

To find the last vestige of Villa Romana de S Cucufate’s factory-scale production requires another half-hour drive to Reguengos’s José de Sousa. Many of its red wines, going back to the 1940s, are legendary. Indeed, it’s still possible to find local restaurants holding bottles from this period, cellared since their release. 

Historically, Alentejo’s wine production has been mostly small scale, for local or family consumption. Only a handful of large producers existed in the 19th and early 20th centuries: Mouchão, Tapada do Chaves, Carmo (now Dona Maria), and José de Sousa. Of these iconic wineries, only José de Sousa has continued with talhas into the 21st century. Getting that far is a tale of death and resurrection.

When death brought dynastic change in the 1970s, talha production was forced on a younger generation with no experience of this technology. Turning to well-meaning but equally clueless modern winemakers resulted in wines that were inconsistent and sometimes undrinkable. Production continued to spiral downward. Sudden death came when three workers died during the 1984 vintage, forcing the winery’s sale. 

Resurrection followed when José Maria da Fonseca purchased José de Sousa in 1985. After taking possession, they soon realized that the cool, dark, old subterranean cellar was in a ruinous state: its 120-talha capacity had dwindled to just 20 vessels, and all of the old wine stock had been sold off.

Part of the original negotiating team, Fonseca’s UC Davis-trained chief winemaker Domingos Soares Franco had a deep respect for its talha production, old wines, and old-vine, Grand Noir-dominant vineyards. If Franco hadn’t been board chairman and a Fonseca family member, there is no way the dilapidated old icon would ever have been restored to its former glory.

The biggest problem was that Alentejo’s once-thriving talha-making industry had died off in the 1920s, so new replacements were simply not possible. Undaunted, Franco and his team set about cleaning and restoring as many surviving talha as possible. After searching far and wide for replacements—antique shops, garden centers, and so on—they amassed around 120 talhas. All were produced between 1819 and 1908, ranged in size from 700 liters to 2,000 liters, and were branded by master craftsmen from S Pedro do Corval (between Reguengos and Monsaraz), Campo Maior, and Vidigueira.

Equally important was restoring lost intellectual capital. A long-retired cellar master from the adega’s golden age was located and brought into the team. His firsthand knowledge proved instrumental in piecing together both common and extraordinary pre-1960s practices. Importantly, a library of more than 100 old wines going back to the 1940s was reassembled through auctions and by swapping new for old with restaurants. Franco wasn’t flying blind now.

The first few vintages were driven by trial and error, with production finally back on track by 1988, and the next few years were spent understanding the technology. One of the first lessons learned was that talhas demand serious respect. 

In the early days, it was not unusual to open the cellar doors on a morning to find that overnight CO2 buildup had turned talhas into potentially lethal bombs. Massive 2in- (5cm-) thick shards of clay were tossed as far as 33ft (10m) away. A few ruined talhas later, and this expensive, dangerous problem was eliminated by more cap-breaking, with punch-downs carried out later into the night.

They also discovered one of talhas’ key advantages over buried qvevri. It turned out that the raised, rope-like decoration around each talha’s neck played a central role, controlling its in-built heat-exchange system. Water, poured from above, twists into rivulets that are evenly distributed across the talha’s belly during fermentation. Further absorption is aided by the clay’s purposefully rough, unfinished surface. 

Whereas fermentation temperatures may rise as high as 104°F (40°C), with regular dampening these can be lowered markedly, to 64–68°F (18–20°C)—well below what’s needed for red-wine production. Over time, Franco’s team have learned to fine-tune this process for each vintage’s requirements. Similarly, the system helps maintain consistent maturation within the adega’s normally cool cellar temperatures during summertime spikes.

All this begs for a comparative carbon-neutral study pegging talhas against less efficient, less consistent, electrically driven, refrigerator-clad, stainless-steel tanks. Bets on which would win?

Assembling the pieces

Another brilliant aspect of talhas is their natural, gravity-fed, filtration system. After fermentation, gravity pulls seeds to the bottom, lees settle atop, with a further layer of stems and skins. Although the initial 20 liters of wine is cloudy when tapped from below, the remaining 1,000+ liters drain clearly thereafter. After returning the first 20 liters to the top, this eventually filters through cleanly as well.

Making large quantities of wine at José de Sousa differs from smaller talha production. The adega has two massive stone lagares for foot-treading, which drain into the talhas at subterranean level. This is normally a northern Portuguese technology, but the larger iconic producers mentioned above had a similar setup, sometimes in marble, other times granite.

Lagares function in several ways, offering more winemaking options. Grapes, either whole-cluster or destemmed, are trodden by foot to ensure soft tannin extraction and gentle crushing. Complete crushing is essential in talhas to avoid carbonic maceration and reduce CO2 buildup. Any of what goes into lagares can end up in talha, allowing for finer tuning of tannin management. Additionally, lagares are employed for cold-soak, pre-fermentation maceration. But they can also be used to start fermentation or run a full fermentation. All then moves on to talhas.

Franco has toyed with all of these options and offered me three experimental wines to taste blind. Each was made from Grand Noir grapes (a crossing of Aramond Noir and Petit Bouschet), but fermented differently. The first (foot-trod, then talha-fermented) presented soft, savory dried fruit, mushroom aromas and flavors, with a touch of spice. Complete, nicely integrated, and continuous in the mouth, it was structured with clean, firm tannins. The second wine was fermented only in lagar, then moved to talhas. It offered very broad floral aromas with strong leafy/stem notes. Rounder and more consolidated texturally, surprisingly, its tannins were much softer. The third, a control fermented in a modern, stainless-steel tank, had chocolate/mocha aromas and peppery florals—halfway between the others aromatically. Considerably less integrated, it had sharper acids and tannins.

One intriguing piece of technology inherited in the chattels was an old wooden destemming box. The retired winemaker indicated that, after separating stems from skins, the stems were crushed again separately and the juice was fermented in a 60-liter talha to concentrate tannins. This was back-blended into the larger pots after fermentation to adjust tannins—an early, more natural form of powered tannin addition. Franco is certain that this technique was used in the legendary reds of the 1940s–60s.

Armed with these tools, Franco has worked steadily toward recreating José de Sousa’s surviving wine styles from the 1940s–60s. From discussions with the old cellar master, these wines appear to have been produced in lagares and talhas in tandem, probably following ferments with concrete and/or large chestnut barrel maturation. Lacking large chestnut barrels early on, Franco turned to small French oak barriques. He now suspects that chestnut is the missing ingredient.

Alongside talhas, another important factor in replicating José de Sousa’s old wine styles was the vines. Looking at his oldest mixed vineyard, it became clear that Grand Noir’s predominance was intentional, with its dark-chocolate, fig, and dried-fruit characters central to the character of José de Sousa’s finest vintages. Over time, new plantings have steadily increased its proportion over the blend’s other varieties, Trincadeira and Aragonez.

After 30 years, Domingos Soares Franco is beginning to feel like he’s finally fitting in the last few pieces of the puzzle that produced José de Sousa’s finest vintages. 

Talha’s resurgence

A few years ago, talhas teetered on the edge of extinction. Fortunately, there has been a significant renaissance, and the future looks much brighter. While the rise of international styles and globalization had driven many talha producers into the ground, the recent rise of “natural” wine has opened new doors. Wine culture generally is now more tolerant, more open-minded.

Head of Alentejo CVR’s certification department Luis Amorim comments, “Traditional talha wine was made to drink before summer. There was no money to buy sulfur dioxide, so the wine went to vinegar in the heat. Sulfur-free winemaking is central to the tradition.” 

This approach fits in with the rise of natural wine bars in English-speaking countries, where wine is served by the glass from small, hand-pumped, stainless-steel barrels fresh from wineries. Because talha wine needs colder cellar temperatures and to be drunk young, the logic of local tascas suddenly makes more sense—which is quite ironic, considering they invented the concept in the first place!

Perhaps the most important development has been the official DOC designation in 2010 for talha-made wine. Lobbying for this began with a core group of 15 allied producers from Friar, co-branding as Vitifrades. Now the greater movement stretches to the far edges of Alentejo’s boundaries with eight official Talha DOC regions: Portalegre, Borba, Evora, Redondo, Vidigueira, Granja/Amareleja, Emora, Moura, and Reguengos. The number of DOC wines doubled over 2013–14. Luis Pedro Amorim now spends “two full days of clinical tasting to certify talha wines, where before this took a few hours.”

Alentejo CVR’s former president, Dora Simoes, elaborates: “The vast majority of talha wines are made by independent producers and are not widely distributed. The talha certification is quite new in the region, and many small producers—though they vinify in talha and do it all by the book—at the end of the line, they do not send their wines to be certified simply because they sell it mostly locally and many times do not bottle them.” Simoes thinks “we should be focused on maintaining the tradition of talha locally, on making sure that the producers maintain it as part of the history of the region and follow the traditional techniques of talha winemaking.”

The central truth is that talha wines were created to be drunk fresh in a local tasca with local food. Bottling is a modern afterthought. People need to travel to talha wines to appreciate their essential qualities, rather than the other way around.

All things considered, there has been a resurgence of pride in talhas on many levels. The village of Cabeção, for example, is reported to have 200 families making wine in talhas now. The Vitifrades group hosts a local competition in Vila de Frades in Vidigueira during its annual Talha Wine Festival in early December. Other villages are now holding similar events to see who is making the best wine each year.

Talhas are no longer seen as something shamefully old- fashioned, to be hidden away. They are becoming as hip as they are fun. Grandfathers are eagerly supported by their grandchildren, as whole families make wine together again.

The next wave

Down in Alentejo’s southeast corner, near the Spanish border, I met with a local group of professional talha winemakers in Amarelaja. We tasted together in a newly opened tasca, Adega do Piteira, sitting among a half-dozen talhas ready for the next vintage. The wines I tasted from Cooperative de Granja-Amareleja and Piteira’s own no-sulfur Talha DOC wines were pretty smart. As retired Professor Virgilio Loureiro rightly observed, talha’s former amateur days “of high VA and oxidation are over.”

As in central Alentejo, the talha tradition here is for white grapes: Rouperio, prized for its special honeyed aromas (called “toasty” notes by locals); an ancient grape variety, Perrum (offering structure, tannin, and acidity); and Rabo de Ovelha, for quantity.

I noted that Georgian qvevri wines were mostly white as well, and Loureiro offered an intriguing theory as to why. His research indicates that medieval Cistercian monks had exclusive rights to “tint” wine. This entailed adding specially macerated dark red wine to whites to match the color of blood for sacrament—hence the Iberian terms for red wine, tinto/ tinta. No one was allowed to drink red wines, which were called black wines and considered to belong to the devil. The only safe option for common people was to make and drink white wine. Intriguingly, I was told in Georgia—a deeply Christian culture—that red wine was traditionally called “black wine” there, too.

Moving on to Alentejo’s northeast corner, Terrenus winemaker Rui Reguinga is a good example of conventional winemakers crossing over to talha production. With one talha vintage behind him, he is convinced talhas allow you to feel the strength and purity of a wine more clearly: “You find more minerality and freshness. You get the grapes as they are. You feel the greenness if unripe, which I like, because they aren’t tempered by barrels. It shows the terroir and season much more.” A staunch champion of the cooler region of Portalegre, he continues, “Until the 1970s, most wine was made in talhas.” He mourns the sharp decline in production since. Previously around 400ha (1,000 acres), it’s now only 46ha (113 acres).

Returning to Vidigueira, Heredade do Rocim’s ultra-modern winery is wrapped inside beautiful minimalist architecture. Hidden behind this is the family’s original adega. There, a new generation is restoring Rocim’s tradition of talha winemaking. 

Pedro Pegas, Rocim’s viticulturist, makes this wine from the quinta’s old field-blend vineyard. This was replanted 60 years ago to its previous mix: Aragonez 50%, Trincadeira 30%, Moreto 10%, and Tinta Grossa 10% for reds; Antão Vaz 40%, Perrum 20%, Rabo de Ovelha 20%, and Manteudo 20% for whites. Originally from Vidigueira, Pegas learned the art of talha from his 80-year- old father, who made wine to sell locally. Both recall that Rocim’s old talha wines were highly respected throughout the region.

Both red and white are unsulfured, made with whole clusters, all stems and skins left to ferment and macerate for two months. The wines are bottled for St Martin’s Day, and when I tasted them eight months later, they were elegant, pure, and looking remarkably modern. With an ironic wink, Pedro said that “traditionally there was more variability in how tannins were managed, by adjusting skin contact.” Alas, because current DOC regulations require photographic evidence of grapes in the talha on certain dates, this is no longer possible.

Talhas, our newest technology

Juxtaposed with modern winemaking practices, terracotta- produced wines clearly demonstrate that ancient technology isn’t primitive; it’s merely different. As such, this old technology is really our newest. What it deserves is a greater scientific understanding of its transformational processing, as well as a serious exploration of the new places it can take wine. 

It can spin us forward through the past.

Video of Talha heroes describing their tradition here.