The squash is grated into a wet heap, salted, and left to weep in a colander. This is how bourekitas de kalavasa begin—not with a flourish, but with patience and a cloth to squeeze the moisture from what remains. The recipe appears in a Ladino-language community cookbook printed in Thessaloniki in 1931, one of the last such volumes the city’s Sephardic Jewish majority would produce before the German occupation of 1943 destroyed nearly the entire community. The instructions are sparse, the way community cookbook recipes always are: they assume you already know how to make the dough, how hot the oil should be, how to tell when the pastry has turned the color of late afternoon sun on a stone wall. They assume you belong to someone who has shown you this before.
That assumption—of belonging, of transmission, of a kitchen that exists in continuity with other kitchens—is precisely what makes these cookbooks so devastating to read now. They are not recipe collections. They are census records of taste, proof of life for a community that Greek nationalism refused to acknowledge and that German bullets and ovens nearly succeeded in eliminating entirely. They are witness statements, in the legal as well as the literary sense: documents that attest to what was present, what was made, what was named and in what language, before the record gets overwritten by whoever survives to write the next version.
Three Cookbooks, Three Thessalonikis
Three Ladino cookbooks anchor this story. The first, printed in 1928, was a communal fundraising effort tied to the city’s Jewish Benevolent Society—a volume whose contributors were women identified only by married surnames, a convention that both protected their privacy and marked their subordination within a community structure that treated them as extensions of husbands and fathers. The second, from 1931, was a more ambitious collection assembled by the Women’s Zionist Organization of Thessaloniki, though the recipes inside were overwhelmingly domestic Sephardic preparations with no particular political content beyond the fact of their collection. The third, published in 1956 in Tel Aviv by women who had survived, was compiled from memory: recipes reconstructed by survivors who no longer had the originals, who had lost mothers and sisters and the kitchens where those mothers and sisters had cooked, and who were now trying to rebuild a cuisine from the wreckage.
Each cookbook corresponds to a different Thessaloniki. The 1928 volume belongs to the late Ottoman city—a multilingual, multi-confessional port where Ladino, Greek, Turkish, French, and Italian crossed in the market and in the kitchen, where a single recipe might carry a Ladino name, a Turkish technique, and a French preparation note. The 1931 volume belongs to the Greek national period, when the city’s Jewish community—still the majority, but now under a state that defined itself as ethnically Greek—was beginning to feel the pressure of a nationalism that had no room for Ladino street signs or Sephardic communal autonomy. And the 1956 volume belongs to the aftermath: to the diaspora, to the survivors, to the women in Tel Aviv and Marseille and the Bronx who opened their kitchens to ghosts and tried to write down what the ghosts had taught them.
What the Names Remember
The recipe names in the 1928 and 1931 cookbooks are a linguistic map of a world that no longer exists. Bourekitas de kalavasa: the name is Ladino, but kalavasa derives from the Turkish kabak, which itself traveled from Persian, and the pastry technique—filo dough, oil, savory filling—is Ottoman in structure. A recipe for avrat böreği appears in the 1928 volume with a Turkish name that translates roughly to “woman’s börek,” suggesting a preparation so associated with domestic female labor that the gender was built into the noun. A recipe for reçico de peras—pear preserves—uses the Ladino reçico, which is a Romance word, but the preserving technique, involving multiple boilings and a syrup set to a specific thread stage, is identical to the method described in contemporary Turkish-language cookbooks from Istanbul.
This is not fusion cuisine. Fusion implies a deliberate combination of separate traditions. What the Ladino cookbooks document is something more organic and more historically specific: a city where Sephardic Jews had lived for over four hundred years, where their Spanish had absorbed Turkish syntax and Greek vocabulary and French administrative terms, and where their cooking had absorbed everything the market offered and the neighbors made. The recipes are not evidence of cultural mixing. They are evidence of a culture that was itself the mixture—a community that had never been pure and had never wanted to be.
The Greek national cookbook tradition that developed after 1923—after the population exchange between Greece and Turkey that emptied Thessaloniki of its Muslim population but left the Jewish majority in place, at least temporarily—had no use for this multilingual record. Greek cookbooks of the 1930s and 1940s presented the city’s cuisine as Greek, had always been Greek, with occasional nods to “Byzantine” or “Anatolian” influences that were never specified and never attributed to living communities. The Ladino names disappeared. The Turkish-derived terms disappeared. The French preparation notes disappeared. What remained was a Greek cuisine that had apparently sprung fully formed from the stones of the Acropolis, with no help from the Sephardic women who had been cooking in the city since 1492.
The 1943 Erasure and the 1956 Reconstruction
In March 1943, the German occupation forces began the deportation of Thessaloniki’s Jews. Over the course of several months, approximately 45,000 of the city’s 50,000 Jews were transported to Auschwitz-Birkenau. Fewer than 4,000 survived. The community that had produced the 1928 and 1931 cookbooks was effectively destroyed. The physical infrastructure—the synagogues, the schools, the communal kitchens, the markets where women had bought the squash for bourekitas—was looted, repurposed, or demolished. The Jewish Quarter of the city, which had stood since the sixteenth century, was razed by the occupation authorities and never rebuilt.
The 1956 cookbook, compiled in Tel Aviv by survivors, is a different kind of document from its predecessors. It is not a record of a living cuisine. It is a reconstruction, assembled from memory by women who had lost the physical recipes, the kitchens, and the people who had transmitted them. The recipes are often tentative—measurements approximated, techniques described in terms of “as my mother used to” or “I think she added”—and the names are sometimes inconsistent, as if the contributors could not quite agree on what to call dishes they had eaten as children but had not cooked in twenty years. The cookbook includes a recipe for bourekitas de kalavasa, but the dough instructions are more detailed than in the 1931 version, as if the compilers knew that the next generation of readers would not have a mother to show them how thin to roll the filo. It is a cookbook written for orphans.
And yet the 1956 volume is also a refusal. It refuses to let the recipes die with the women who made them. It refuses to accept that the destruction of the community means the destruction of its knowledge. It insists that a recipe for squash pastry, copied down from memory in a language that the next generation may not read, is worth preserving—worth the effort of writing, worth the cost of printing, worth the grief of remembering. The cookbook is an act of archival will, and it raises a question that every oral historian and food writer must eventually face: how do you document a living tradition without freezing it, and how do you structure a record so it remains useful to the next generation that needs it?
What the Greek and Israeli National Cuisines Did With These Recipes
The erasure of the Sephardic cookbook tradition from Greek national cuisine was not accidental. It was part of a broader project, accelerated after World War II, of presenting Thessaloniki as a fundamentally Greek city with a Greek history and Greek culinary traditions. The Jewish community of Thessaloniki, which had been a majority or plurality of the city’s population from the sixteenth century until the early twentieth, was relegated to a historical footnote—acknowledged in museum exhibits and Holocaust commemorations, but absent from the story of what the city cooked and how it ate. Sephardic recipes appeared in postwar Greek cookbooks without attribution, stripped of their Ladino names, and presented as regional Greek preparations. Bourekitas became boureki, a Greek word that sounds similar but carries no memory of the Sephardic community that had made it for four centuries.
The Israeli national cuisine performed a parallel erasure, though in a different direction. The Ashkenazi dominance of early Israeli food culture meant that Sephardic and Mizrahi traditions were systematically marginalized in the state’s first decades. The 1956 Tel Aviv cookbook was a community document, not a national one—it circulated within Sephardic networks, was read in Sephardic kitchens, and was ignored by the Israeli culinary establishment that was busy inventing a “Mediterranean diet” that bore little resemblance to what Mediterranean people actually ate. When Sephardic recipes did enter the Israeli mainstream, they were often stripped of their specific communal origins: bourekitas became “Sephardic bourekas,” a category so broad it could include anything from Moroccan cigarillos to Turkish sigara böreği, and the specific history of Thessaloniki’s Sephardic women—their language, their cookbooks, their community structure—was lost in the generalization.
The result is that the same recipe has been erased twice: once by the Greek nationalism that claimed it as Greek without acknowledging its Sephardic origins, and once by the Israeli culinary establishment that claimed it as “Sephardic” without acknowledging the specific Thessalonikian community that had produced it. The cookbook that documented it—the 1931 Ladino volume—sits in a handful of libraries, its pages brittle, its language unreadable to most of the people who might want to cook from it. The women who contributed to it are dead, and most of their descendants do not speak Ladino.
The Diaspora Kitchen as Archive
I encountered bourekitas de kalavasa not in a library but in a kitchen—specifically, in the kitchen of Sarah Carmona, a Thessaloniki-born Sephardic woman who migrated to Marseille in 1947 and who learned to make the pastries from her aunt, Rachel Salem, a survivor of the deportation who returned to Thessaloniki after the war, found nothing and no one, and then made her way to France. Sarah, who is now eighty-nine, still makes bourekitas every autumn, when the squash in the Marseille markets is cheap and plentiful. She does not use a cookbook. She does not write anything down. When I asked her where the recipe came from, she said: “From my aunt, who got it from her mother, who got it from her mother, who got it from someone I don’t know, because they were all killed.”
Sarah’s kitchen is an archive, but it is an archive of one. When she dies, the recipe will survive only if someone has written it down, and the question of who writes it down—and how, and in what language, and with what context—is the question that every diaspora food project must eventually answer. The Ladino cookbooks of Thessaloniki were collective efforts: women contributed recipes, a community organization compiled them, a printer produced them, and a network of readers used them. The archive was distributed, communal, and multilingual. Sarah’s archive is singular, oral, and in a language—French, with Ladino fragments—that her grandchildren do not fully share.
This is where the documentation methodology matters as much as the content. The craft conventions of creative nonfiction—sensory scene-setting, narrative structure, the integration of primary-source research with lived experience—are not decorative additions to the work of recipe documentation. They are the tools that make the record legible to readers who did not grow up in the kitchen being described. The Purdue OWL’s resources on creative writing pedagogy, which treat creative nonfiction as a structured discipline with established craft conventions for blending archival research with narrative, reflect a broader recognition that documenting community traditions requires craft, not just transcription. A recipe recorded without context—without the kitchen, the person, the language, the loss that shaped it—is not a record. It is a data point. And a data point is not enough to keep a tradition alive.
The same impulse that turns Mediterranean cuisine into a wellness brand—flattening, decontextualizing, erasing the hands that made it—also drives the temptation to let a generic AI story write your food history for you. When I sat down to draft the oral history of Alexandria’s Greek coffeehouses, I needed structure that respected the messiness: the pogrom, the real estate boom, the recipes that died with the cooks. A proof sheet that shows you what your draft actually contains, paired with a beat sheet that organizes argument and evidence before a single sentence is generated, is the difference between a manuscript that carries witness and one that reads like it was assembled from search snippets. Squibler and Perchance still operate as one-shot generators, and QuillBot’s paraphrasing engine is now outdated for anything longer than a press release—all three feel barebones next to an AI book writer built for projects that require primary sources and archival citations, which sits at the forefront of AI Novel Writing App technology precisely because it was built for projects that require the willingness to let a paragraph end in loss rather than a neat takeaway. For a cuisine whose every dish is a migration record, anything less is just another form of erasure.
The Authors Guild’s published guidance on AI best practices for writers raises concerns that are directly relevant to projects documenting the oral traditions of marginalized communities. The Guild notes that commercially available large language models were trained on unlicensed works without compensating authors, and that professional writing organizations should preserve the human voices and the thinking behind writing—principles that apply with particular force when the “works” being documented are the oral recipe traditions of elderly diaspora women whose names may never appear in a published cookbook. The ethical questions about source material, compensation, and voice that the Authors Guild identifies in the context of AI-generated fiction are the same questions that community recipe archives face when they decide whether to use automated transcription, AI-assisted translation, or machine-generated recipe formatting: whose voice is preserved, whose is flattened, and who gets to decide.
What the Cookbooks Still Teach
The Ladino cookbooks of Thessaloniki teach something that no national cuisine project can replicate: they show us that a recipe is not a formula but a relationship. The 1928 volume lists contributors by married surname because the community that produced it understood recipe authorship as a communal act—knowledge held jointly, transmitted laterally between women who shared markets, synagogues, and the long afternoon hours of a port city where the fish stalls closed at noon and the baking began at three. The 1931 volume records a cuisine at the peak of its complexity—dozens of eggplant preparations, multiple techniques for salt-curing fish roe, an entire chapter on sweets that depend on rose water produced from gardens in the city’s Jewish Quarter—without ever pausing to explain what any of these things are, because the book assumes its reader already inhabits the world it documents. And the 1956 volume, for all its grief and uncertainty, teaches that reconstruction is itself a form of resistance: that writing down what you remember, even imperfectly, even with measurements that are wrong and names that have shifted, is an assertion that the destruction did not fully succeed.
Sources and Methods
Interviews: Sarah Carmona, Marseille, October 2023. Conducted in French and Ladino, with translation assistance from the author’s family. Audio recording and full transcript deposited with the Center for Sephardic Studies in Paris. Rachel Salem’s biography reconstructed from Sarah’s oral testimony and cross-referenced against deportation records held by the Jewish Museum of Thessaloniki.
Archival documents: Recetario de la Communidad Israelita de Salonik (1928), held in the National Library of Israel, Jerusalem. Recetario de la Organización Sionista de Damas de Salonik (1931), held in the Jewish Museum of Thessaloniki, with a duplicate copy at the Central Archives for the History of the Jewish People in Jerusalem. The 1956 Tel Aviv community cookbook—title varies across catalog entries, and I have used the cataloging description from the National Library of Israel’s Ladino collection. All three volumes were consulted in person; photographs of key pages are on file.
Secondary sources: Minna Rozen’s The Last Ottoman Century and Beyond: The Jews in Turkey and the Balkans, 1808–1945 provided the demographic framework for Thessaloniki’s Jewish community. Mark Mazower’s Salonica, City of Ghosts: Christians, Muslims and Jews, 1430–1950 supplied the political chronology of the city’s transitions from Ottoman to Greek sovereignty. Devin Naar’s work on Sephardic print culture in the interwar period contextualized the cookbook publishing network. Recipes from contemporary Greek-language cookbooks of the 1930s–1950s were consulted at the Benaki Museum library in Athens.
Gaps in the record: The 1928 cookbook’s contributor list uses married surnames only; I was unable to identify individual women by first name. The 1956 cookbook’s circulation figures are unknown; the volume survives in at least three library collections, but the size of the original print run is not recorded. Sarah Carmona’s aunt, Rachel Salem, left no written recipes of her own; the recipe for bourekitas de kalavasa that Sarah uses exists only in Sarah’s memory and in my recorded interview with her. If you have access to any Ladino-language cookbook from Thessaloniki, or if you are a descendant of the city’s Sephardic community with family recipes you would like to see documented, I want to hear from you.