Nearly every Skellig unit eventually arrives at some version of the question of what Skellig actually is, whether he is meant to be read as a literal angel, an owl-like creature, an elderly man in physical decline, or some deliberately unresolved combination of all three. This is genuinely the novel's central interpretive puzzle, and Almond has been explicit in interviews that he intended the ambiguity to remain unresolved rather than functioning as a mystery with a hidden correct answer. Teachers grading essays on this question need a rubric built from the start around this authorial intent, rather than one that implicitly rewards students for guessing whatever interpretation the teacher personally favors.

The most common grading mistake on this prompt involves teachers unconsciously scoring essays higher when they happen to align with the teacher's own preferred reading of the character, which is an understandable but genuinely unfair bias given how emotionally engaging this particular interpretive question tends to be. A student who argues persuasively that Skellig is best read as a literal angel deserves the same grade as an equally persuasive student arguing he represents Michael's psychological coping mechanism, provided both essays meet the same standards for evidence and argument quality.

Building the rubric around argument quality rather than conclusion also protects against a subtler bias, where teachers unconsciously reward the most creative or unexpected interpretation over a more straightforward but equally well-supported one. A conventional, well-argued reading of Skellig as simply mysterious and unexplainable deserves full credit alongside a more inventive psychological or symbolic reading, as long as both are grounded in specific, accurate textual evidence rather than speculation disconnected from the actual novel.

Structuring a Bias-Resistant Rubric

A practical way to reduce grading bias on this prompt is to draft the rubric using several different sample student theses in advance, checking that each one could plausibly earn full marks if equally well supported, before ever seeing actual student submissions. This exercise forces the rubric author to notice and correct language that implicitly favors one interpretation over another, such as criteria that only make sense if Skellig is assumed to be supernatural rather than psychological, and helps ensure the final rubric genuinely accommodates the full range of legitimate readings the text supports.

• A clearly stated, specific claim about Skellig's nature or symbolic function
• Evidence drawn from at least three distinct scenes across the novel
• Acknowledgment of at least one piece of complicating or ambiguous evidence
• Consistency between the claim and the evidence used throughout the essay
• A conclusion that reflects on why the ambiguity itself might matter

Handling the Complicating Evidence Requirement

Requiring students to acknowledge at least one piece of complicating evidence, meaning a textual detail that does not fit neatly into their chosen interpretation, is one of the more valuable and challenging elements of this assignment, since the novel genuinely contains details that resist every single interpretation to some degree. A student arguing Skellig is purely symbolic still has to account for physically observable details like his need for food and his gradual physical healing, and requiring them to address this directly produces more intellectually honest and more sophisticated essays than a rubric that allows students to simply ignore inconvenient evidence.

Grading this criterion fairly requires teachers to resist the temptation to demand that students fully resolve the complicating evidence, since the novel itself never resolves it either. The bar should be genuine acknowledgment and thoughtful engagement, not resolution, and essays that end with something like this detail remains genuinely puzzling and may be part of the novel's larger point about accepting uncertainty should be scored as successfully meeting this criterion rather than penalized for failing to tie up every loose end.

Supporting Consistent, Bias-Aware Grading at Scale

Because this prompt is so emotionally engaging and so prone to unconscious grading bias, it is one of the clearer cases where a structured, rubric-anchored grading tool can genuinely improve fairness across a large class set, applying the same criteria regardless of which specific interpretation a student happens to argue for. AI-assisted grading configured strictly around the rubric's stated criteria, rather than around any single expected answer, helps catch the kind of unconscious scoring drift that even experienced, well-intentioned teachers can fall into when grading dozens of essays on a topic they personally find fascinating.

This kind of support works best as a check on consistency rather than a replacement for a teacher's own literary judgment, flagging cases where two similarly argued essays received noticeably different scores so a teacher can review and reconcile the discrepancy before final grades are recorded. For a prompt this central to the unit and this prone to personal interpretation, that extra layer of consistency checking meaningfully strengthens student trust that their grade reflects the quality of their thinking rather than whether they happened to agree with their teacher.