You likely can't select text in an academic paper PDF because it's image-based rather than text-based — meaning the document was created by scanning physical pages without optical character recognition (OCR), or the text layer is corrupted or missing.
Common Causes
Scanned PDFs without OCR: Many older papers, conference proceedings, and archival documents were digitized by scanning printed pages. Without OCR processing, these files contain only images of text, not selectable characters.
Security restrictions: Some publishers apply copy-protection settings that disable text selection to prevent unauthorized redistribution. You'll typically see a "not allowed" cursor when hovering over text.
Corrupted text layer: Occasionally, PDFs generated from certain typesetting systems or converted from other formats end up with damaged or invisible text layers, even though the visual text appears normal.
Embedded fonts issues: If a PDF uses custom or improperly embedded fonts, your PDF reader may fail to recognize the characters as selectable text.
How to Fix It
Run OCR software: Tools like Adobe Acrobat Pro, ABBYY FineReader, or free options like OCRmyPDF can add a searchable text layer to scanned documents. This process converts the images into machine-readable text.
Check document permissions: Right-click the PDF and view properties or security settings. If copying is disabled, you may need to contact the publisher for an accessible version or use your institution's library access, which sometimes provides less-restricted copies.
Try a different PDF reader: Sometimes switching from Preview (Mac) to Adobe Reader, or from one browser's built-in viewer to another, resolves rendering issues.
Re-download from the source: If you suspect file corruption, obtain a fresh copy from the publisher's website or your university library portal.
For papers you frequently reference, consider uploading them to WisPaper's My Library, which automatically extracts metadata and lets you chat with your documents even when the original text layer is problematic — the AI can still parse and interpret the content for you.
