Searchable layers
A concordance gathers occurrences matching a query and displays context. Decide whether you are searching exact transliteration, normalized spelling, lemma, translation, morphological tag or dictionary text. These layers answer different questions. ETCSL provides a glossary and search interface for a defined Sumerian literary corpus; ePSD2 offers Sumerian glossaries and lemmatized texts across linked Oracc projects. TLA supplies a corpus and lemma lists for Egyptian. Their results are shaped by which texts and editorial layers each project includes.
For every search, preserve the query string, filters, corpus name, version or access date, and number of hits. Open the texts behind representative results. A concordance line can hide damage, duplicate manuscripts, variant editions or a quotation embedded in a later text. The object or critical edition remains the place to check those details.
Frequency and absence
Raw hit counts are not frequencies of speech. A corpus may be dominated by administrative records, temple texts or copied literature. It may include many witnesses to one composition; a large number of tokens can reflect preservation rather than historical prevalence. A useful frequency claim needs a denominator, sampling frame and policy for duplicates and restorations. Compare like with like: words per thousand running tokens in comparable dated subsets, if the data support that measure.
No match in one database means only that the query found no matching indexed item there. It may reflect spelling, OCR, transliteration, lemmatization, publication gaps or damaged witnesses. Before arguing that a word is absent from an ancient period, search variants and other corpora, check the database’s coverage and record the remaining limit.
Collocation
A collocation is an observed tendency for words to occur together. To test one, collect occurrences of a lemma with a defined context window or syntactic relation and compare them with a baseline. A striking pair in two parallel copies may be one inherited formula rather than two independent uses. For interpretation, inspect complete constructions: a preposition, case frame or object class can be more informative than neighboring words alone.
Practice
Search a lemma in a corpus with a published scope. Sample ten hits or all hits if fewer. Record date, genre, construction, edition and whether each hit is secure or restored. Group recurring constructions and write a paragraph on what the sample does and does not establish about meaning.
Modern-language examples on this page clarify a linguistic pattern. They do not establish unattested sounds or forms for an ancient language. Follow the cited sources below and keep editorial readings separate from the surviving trace.
