Trovadorismo Medieval

How to work with trovadorismo medieval source material without losing your mind

Most people approaching the Cantigas Galego-Portuguesas start by reading a modern edition and assuming the text is settled. It isn't. The three main codices — the Cancioneiro da Ajuda, the Cancioneiro da Biblioteca Nacional de Portugal (formerly Colocci-Brancuti), and the Cancioneiro da Vaticana — were copied by different scribes at different times, often from lost exemplars, and each contains variant readings that change the meaning of entire stanzas. I learned this the hard way when I was building a comparative edition of the cantigas de amor and spent three weeks tracking what I thought was a single textual tradition, only to discover the Ajuda and Vaticana versions of a Dom Dinis cantiga diverged on every second line. The standard classification divides the corpus into cantigas de amigo, cantigas de amor, and cantigas de escárnio e maldizer. That classification is useful but misleading if you treat it as a rigid system. The scribes who compiled these codices didn't organize them that way. The Ajuda manuscrito arranges songs by melody type and tonal pattern, not by genre. The Vaticana uses a different logic altogether. What we call "genres" today are largely a scholarly construction imposed on a much messier original reality. Some cantigas share features across categories. A cantiga de amor might use the first-person female voice typical of amigo poetry. That's not an error — it's how the tradition actually worked. For anyone trying to engage with the material seriously, the real starting point is understanding the manuscript tradition, not jumping straight to interpretation. The vocabulary itself requires a special kind of attention. Medieval Galician-Portuguese contains words that have disappeared from modern usage entirely, and words that look familiar but carry different meanings. The word "amigo," for instance, doesn't simply mean "friend" in the modern sense. In the context of cantigas de amigo, it refers to a romantic partner, usually male, from the perspective of a female speaker. Confusing this with contemporary usage will distort your reading of half the corpus.

I ran into a specific problem with a Pedro Garcia de Guilhade cantiga where the editorial convention of the time had normalized a word that didn't actually appear in any of the three main codices. The editor had supplied the missing term based on parallel passages in other songs. That's standard editorial practice, but it means the text you're analyzing isn't purely the medieval source. It's a reconstruction. I flagged this in my notes and cross-referenced every supplied reading against the facsimile images. It took longer but it matters if you're doing citation work or building a critical edition.

Practical workflow for studying the Trovadorismo Medieval corpus

Start with the facsimile. The digital versions from the Biblioteca Nacional de Portugal and the Vatican Library are freely available. Reading a modern critical edition first will anchor you to one particular textual choice and make it harder to notice when another manuscript disagrees. Open the facsimile alongside whatever edition you're using. The scribe's handwriting in the Ajuda is relatively clear compared to the Vaticana, where some folios are badly damaged. The Colocci-Brancuti sits somewhere in between. Build a simple comparison table for each cantiga you work on. Columns for each manuscript, rows for each line or stanza. Note where they agree and where they diverge. Even a basic spreadsheet makes patterns visible. You'll start noticing that certain variants cluster in specific codices, which points to separate recensions or scribe traditions rather than random error. When you hit a word you can't parse, check the Dicionário do Português Medieval first, then the Vocabulário do Português Medieval by Cândido Luna. These are the standard reference tools and they cover the vast majority of problematic terms. For proper names and toponyms, the works of José Carlos Santos and António Resende de Oliveira on onomastics in the cantigas are essential. Don't skip the phonological and morphological notes in these references either. Medieval Galician-Portuguese spelling was largely phonetic, so a single word might appear with five different spellings across the manuscripts, all representing the same underlying form.

Common mistakes and where the material breaks down

The biggest pitfall is treating the trovadores as a monolithic literary movement. They weren't. The corpus spans roughly two centuries, from the late 12th to the late 14th. The poetic conventions, musical structures, and even the social context shifted significantly over that time. Martín Codax, writing in the late 13th century, operates in a different cultural milieu than the earliest anonymous cantigas. Comparing them as if they're part of the same project misses important developments. Another issue is the assumption that the three codices capture the full output of the trovadorismo. They don't. The corpus contains roughly 1,680 surviving cantigas, but that's the result of centuries of copying, loss, and selection. We have no way of knowing how much was lost. The fact that we have more cantigas de escárnio and maldizer than one might expect suggests these were particularly durable in oral transmission, not that they were the dominant form. The absence of other types of poetry in the manuscripts doesn't mean they didn't exist. The musical notation is another area where certainty is limited. The Ajuda and Vaticana contain neumatic notation, but it's incomplete and its rhythmic interpretation remains debated. If you're working with the musical dimension, the editions by Alfred E. Dell are the standard reference, but read them with the understanding that the rhythmic values assigned to the neumes are hypotheses, not facts. The Colocci-Brancuti has no musical notation at all, which means those songs survive only as text.

For accessing the source material, the Biblioteca Nacional de Portugal offers high-resolution facsimiles of the Cancioneiro da Biblioteca Nacional and the Cancioneiro da Ajuda. The Vatican Library has digitized the Cancioneiro da Vaticana. The project "Cantigas de Amigo" at the Universidade de A Coruña also provides searchable transcriptions with manuscript sigla. None of these are perfect, but they're the best you'll get without doing paleographic work yourself. The trovadorismo medieval yields to careful study, but it resists easy synthesis. The sources are fragmentary, the editorial decisions are rarely transparent, and the gap between the written codices and the oral performance tradition they tried to capture can't really be bridged. The best approach is to work directly from the manuscripts, keep your editorial dependencies visible, and accept that some questions won't have definitive answers.