Deep Dive: Reciprocal and Inverse Trigonometric Functions

Provenance (2026-09-29): read/checked in full for this pass — research/veridian/WMA13-verified-facts.md (grepped and read for every mention of "reciprocal", "sec", "cosec", "cot", "arcsin", "arccos", "arctan" across the whole file — §1's verbatim spec 2.1–2.2 table, §3's formula-booklet/memorise-list split, §4's mark-scheme conventions, the full §5 write-up for 2.1–2.2 including the Jan 2022 Q2 full re-extraction, the Jan 2025 Q1(a) and Oct 2020 Q8 entries, and the §9 lesson-audit-log entry for this exact lesson); content/veridian/curriculum/wma13/reciprocal-and-inverse-trigonometric-functions.ts (the full 1,122-line lesson, read start to end, specifically to identify every definition, derivation, worked example, trap and MCQ it already contains and avoid duplicating any of them); content/veridian/technique-hub/wma13/codex-formula-and-method-bank.md, forge-exam-technique-guide.md and signal-frequency-and-boundaries.md (read/grepped in full for this topic's existing coverage); content/veridian/technique-hub/wma13/codex-deep-dive-differentiation-rules.md and codex-deep-dive-integration-by-recognition.md (grepped specifically to confirm two cross-topic real anchors — Oct 2020 Q8(ii)'s arcsin-beyond-spec finding, and Jun 2023 Q9(c)'s cosec-integration correction — are already fully documented there, so this document points at them rather than reworking them); content/veridian/question-bank/wma13.ts (grepped — confirmed it contains no reciprocal- or inverse-trig item). No new primary-source PDF was opened for this pass; every citation below traces to one of the documents just named. Every micro-example's arithmetic was independently recomputed with Python (sympy and direct math evaluation) before being written in, not carried over from any file on trust. Per this document's own standing rule, mark-scheme and examiner-report wording is paraphrased throughout, with only short (under-15-word) fragments quoted and attributed — the full verbatim text lives in WMA13-verified-facts.md and is cited by location, not reproduced here. Not affiliated with or endorsed by Pearson Edexcel.

#CODEX Deep-Dive: Reciprocal and Inverse Trigonometric Functions (WMA13, spec 2.1–2.2)

#0. What this document is, and what it deliberately doesn't repeat

The lesson file already teaches this topic thoroughly: the three reciprocal definitions and the cosec/sec naming trap; the boundedness and asymptote reasoning ($|\sec\theta|,|\text{cosec}\,\theta|\geqslant1$; asymptotes at the original ratio's zeros); a full concept-ladder and mechanism derivation of $\sec^2\theta\equiv1+\tan^2\theta$ and $\text{cosec}^2\theta\equiv1+\cot^2\theta$ from $\cos^2\theta+\sin^2\theta\equiv1$; the one-one/many-one reasoning that fixes the restricted domains and graphs of $\arcsin$, $\arccos$ and $\arctan$; a worked chain and chain-drill solving squared-reciprocal equations down to a quadratic in $\tan\theta$ or $\cot\theta$; a method-comparison block; a full marked solution (given-ratio-plus-quadrant, hence-solve) with a common-wrong-path; seven named traps; and seven MCQs. The formula-and-method bank gives a compact definitions/domain table and names the two identity-error findings. FORGE's trap table cites this lesson's own error ids. None of that is reworked here.

What none of those documents do is (1) lay out, once, the complete real mark ladder for the one confirmed standalone exam anchor for spec 2.1–2.2 together with its own independently-checked arithmetic; (2) trace where reciprocal- and inverse-trig knowledge is actually load-bearing inside other topics' own real questions — differentiation, numerical methods, integration — as opposed to being tested for its own sake; (3) state plainly, with sourcing, where this topic's boundary actually sits (what a WMA13 paper has never required, confirmed rather than assumed); and (4) report exactly how thin the standalone-anchor evidence is, rather than letting the topic's real theoretical richness imply an exam-frequency richness the sources don't support. That is this document's job.

#1. Where this sits on the paper

Spec 2.1 (paraphrased from the facts bank's own §1 transcription): knowledge of secant, cosecant and cotangent and of arcsin, arccos and arctan, their relationships to sine, cosine and tangent, and understanding of their graphs and restricted domains, in both degrees and radians. Spec 2.2: knowledge and use of $\sec^2\theta\equiv1+\tan^2\theta$ and $\text{cosec}^2\theta\equiv1+\cot^2\theta$.

The booklet split is stark here. The three reciprocal definitions are not printed as such, though the differentiation and integration tables for $\sec x$, $\text{cosec}\,x$ and $\cot x$ imply them; the two Pythagorean-derived identities in 2.2 are on the spec's own explicit "will not appear in the booklet" list — genuine memorised content. The restricted-domain/range facts for $\arcsin$, $\arccos$, $\arctan$ are not printed anywhere either.

Frequency — the single most important honest fact about this topic. SIGNAL's own frequency table records this spec item as embedded as an opening step inside larger questions in most series reviewed, not separately tallied, with exactly one full standalone worked anchor located across the whole archive (Jan 2022), typically worth 1–3 marks as the accessible opening step of a bigger question. This is not a thin-evidence topic in the sense of having little theory — the theory is completely specified and closed — but it is thin in the sense that matters most for exam preparation: real, fully-transcribed standalone past-paper evidence for spec 2.1–2.2 on its own is exactly one question.

#2. The complete real theory, compactly

Definitions. $\sec\theta=\dfrac{1}{\cos\theta}$, $\text{cosec}\,\theta=\dfrac{1}{\sin\theta}$ (also $\csc\theta$), $\cot\theta=\dfrac{1}{\tan\theta}=\dfrac{\cos\theta}{\sin\theta}$ — reciprocals, nothing else. The one real, confirmed naming trap is reaching for cosine when reciprocating "cosec" (the shared first syllable) instead of sine; the lesson's own mnemonic and the check "a reciprocal pair multiplies to 1" both resolve it, and neither is repeated here.

Domain and boundedness, as direct consequences of the definitions, not separate facts. $\sec\theta$ is undefined wherever $\cos\theta=0$; $\text{cosec}\,\theta$ and $\cot\theta$ are both undefined wherever $\sin\theta=0$ — each reciprocal's asymptotes sit exactly at the original ratio's zeros. Since $|\cos\theta|,|\sin\theta|\leqslant1$ always, $|\sec\theta|,|\text{cosec}\,\theta|\geqslant1$ always; $\cot\theta$ carries no such floor, because it reciprocates $\tan\theta$, which is already unbounded.

The two 2.2 identities, one derivation performed twice. Dividing $\cos^2\theta+\sin^2\theta\equiv1$ through by $\cos^2\theta$ gives $1+\tan^2\theta\equiv\sec^2\theta$; dividing by $\sin^2\theta$ instead gives $1+\cot^2\theta\equiv\text{cosec}^2\theta$. Independently re-checked at a fresh angle not used in the lesson's own concept-ladder or elsewhere in this course's WMA13 material, $\theta=50°$: $\sec^250°=2.420277$ and $1+\tan^250°=2.420277$; $\text{cosec}^250°=1.704088$ and $1+\cot^250°=1.704088$ (six s.f. agreement, both identities, one fresh check each).

Why $\arcsin$, $\arccos$, $\arctan$ need restricted domains, in one sentence. None of $\sin\theta$, $\cos\theta$, $\tan\theta$ is one-one over its full periodic domain, and a many-one function has no well-defined inverse until cut down to a one-one piece; the piece chosen for each is the interval containing $\theta=0$ on which the function is monotonic and covers its whole range. That fixes: $\arcsin$, domain $[-1,1]$, range $-90°\leqslant y\leqslant90°$; $\arccos$, domain $[-1,1]$, range $0°\leqslant y\leqslant180°$; $\arctan$, domain $\mathbb{R}$, range the open interval $-90°<y<90°$ (open because $\tan\theta$ is itself undefined at $\pm90°$, so those two values are never reached, not merely large). A value outside $[-1,1]$ handed to $\arcsin$ or $\arccos$ is not a large angle — it is undefined outright, because $\sin\theta$ and $\cos\theta$ never produce it in the first place.

#3. Every real distinct sub-case, with a short worked example each

A. A direct reciprocal equation, solved without ever touching the squared identities — genuinely not built into the lesson. Every equation-solving block in the lesson (worked chain, chain-drill, method-comparison, marked solution) reduces via $\sec^2\theta\equiv1+\tan^2\theta$ or $\text{cosec}^2\theta\equiv1+\cot^2\theta$ to a quadratic in $\tan\theta$ or $\cot\theta$ — there is no example anywhere of the simpler case, a bare linear reciprocal equation. Micro-example (self-generated, not sourced to any exam paper): solve $2\,\text{cosec}\,\theta=5$ for $0°\leqslant\theta<360°$. Rearranging, $\text{cosec}\,\theta=2.5\Rightarrow\sin\theta=0.4$, giving $\theta=\arcsin(0.4)=23.6°$ (1 d.p.) or $\theta=180°-23.6°=156.4°$ (independently checked: $2/\sin(23.578...°)=5.000$ exactly). The coefficient trap this lesson already documents applies at the very first line — $2\,\text{cosec}\,\theta$ must become $2/\sin\theta$, never $1/(2\sin\theta)$ — and this is the simplest possible setting in which to see why.

B. Given one exact ratio and a quadrant, find the other reciprocal ratios exactly — fully covered by the lesson's marked solution ($\sin\theta=\frac45$, $\theta$ obtuse); a short, fresh check with different numbers. Micro-example (self-generated, fresh $9\text{-}40\text{-}41$ Pythagorean triple, not used elsewhere in this course's WMA13 material): given $\cos\theta=-\frac{9}{41}$ with $\theta$ in the third quadrant ($180°<\theta<270°$, where sine is also negative), $\sin\theta=-\frac{40}{41}$, so $\sec\theta=\frac{1}{\cos\theta}=-\frac{41}{9}$ and $\cot\theta=\frac{\cos\theta}{\sin\theta}=\frac{9}{40}$ (both independently verified numerically). The mechanism — use the quadrant to fix the sign the Pythagorean identity alone cannot decide, then reciprocate — is identical to the lesson's own worked example and is not reworked further here.

C. A squared-reciprocal identity collapsing an equation to a quadratic in $\tan\theta$ or $\cot\theta$ — fully covered by the lesson (worked chain, chain-drill, method-comparison) and by the one real standalone anchor's own transition into double-angle territory (§4 below). Not reworked here.

D. Stating the domain or range of $\arcsin$, $\arccos$ or $\arctan$, or recognising a value outside a domain as undefined rather than large — fully covered by the lesson's mechanism block and MCQs (e.g. $\arcsin(1.5)$). One honest caveat worth restating rather than silently inheriting: the lesson's own "domain-of-the-inverse-left-unstated" trap is sourced to two real questions about inverse functions generally (spec 1.2), not to a real question asking specifically for $\arcsin$/$\arccos$/$\arctan$'s own domain or range. No source reviewed for this pass locates a real WMA13 question that tests exactly that. The failure mode is the same shape either way, but the citation should not be overstated as being about this exact sub-topic.

#4. Where this topic's knowledge actually surfaces across the rest of the paper

This is the genuinely new synthesis this document adds: spec 2.1–2.2 is examined almost entirely as an ingredient inside other questions, not as its own subject, and no single document currently traces that across the paper. Five real, independently-locatable instances, in one place:

(i) The one standalone anchor, complete — Jan 2022 Q2 (5 marks total). The question asks students to show $8\cos\theta=3\,\text{cosec}\,\theta$ reduces to $\sin2\theta=k$ (3 marks), then find the smallest positive solution in degrees to 1 d.p. (2 marks). Part (a)'s opening mark is earned simply for stating or using $\text{cosec}\,\theta=1/\sin\theta$ — genuinely accessible, confirmed by the examiner report as earned even by otherwise weak scripts — followed by a method mark for reaching $\sin2\theta=k$ and an accuracy mark for the exact value $k=\frac34$. Independently recomputed: $\theta=\frac12\arcsin(0.75)=24.295...°\to24.3°$ (1 d.p.), matching the real scheme's stated answer. Three findings this one anchor supports, all sourced to the same mark scheme/report and none fabricated: a coefficient-folding error distinct from the cosec/sec mispairing (writing $3\,\text{cosec}\,\theta$ as $1/(3\sin\theta)$, scored zero for the identity mark); the part-(b) accuracy mark is for the smallest solution "only" — a candidate who correctly finds both $24.3°$ and $180°+24.3°$ but keeps both loses it, the mirror image of the far more common missing-a-solution error; and a distinct arithmetic-recovery slip, reporting the doubled angle $2\times24.295...°=48.6°$ rather than dividing by 2. All three are already built into the lesson's own trap taxonomy or its 2026-09-01 audit fix — cited here as the complete picture in one place, not as new findings.

(ii) A numerical-methods root-location question that cannot be started without evaluating a reciprocal trig function — Jan 2025 Q1(a). $f(x)=2\sec x+6x-3$ on $[0.1,0.2]$ (radians). Independently computed: $f(0.1)=2\sec(0.1)+0.6-3=-0.390$ and $f(0.2)=2\sec(0.2)+1.2-3=0.241$, both matching the real mark scheme's stated values to the precision it requires. This question is already used elsewhere in this course's material — for the sign-change/continuity mark-dependency finding and an iteration-mark-dependency finding — but always from the numerical-methods lesson's own angle, never from the angle this document is tracing: a candidate who cannot correctly evaluate $\sec(0.1)$ as $1/\cos(0.1)$ cannot access this question's opening method mark at all, regardless of how well they understand sign-change reasoning. Spec 2.1's reciprocal definitions are a genuine prerequisite skill here, silently assumed rather than tested.

(iii) A differentiation question requiring the product rule applied to a reciprocal trig factor — Oct 2020 Q8(i), not built into any differentiation-topic document in this course. $g(x)=e^{3x}\sec2x$, differentiated by the product rule (2 marks), with the mark scheme explicitly accepting a quotient-rule rewrite of $\sec2x$ as $1/\cos2x$ as an equally valid alternative route — then used to locate a stationary point (the mark detail for that further stage was not transcribed in any source reviewed for this pass, and is not invented here). Independently differentiated and checked with sympy: $g'(x)=e^{3x}\sec2x(3+2\tan2x)$, using the P3-printed result $\sec x\to\sec x\tan x$ with the chain rule folded in for the "$2x$." This is a real, sourced anchor that neither differentiation-rules.ts nor its own CODEX deep-dive currently uses — both instead build their own $e^{3x}\sin2x$ example for the identical two-chain-rules-into-one-product-rule structure — so it is presented here, once, as the confirmed real instance of that structure that happens to feature this topic's own function.

(iv) A recognition-integration question resting on the reciprocal-trig differentiation table — Jun 2023 Q9(c) — documented and corrected in full elsewhere; pointed at, not reworked. codex-deep-dive-integration-by-recognition.md (§3.5) independently reconfirms a correction already made in the formula bank: this question tests $\int\text{cosec}\,x\cot x\,dx$, read backwards off the P3-printed derivative $\text{cosec}\,x\to-\text{cosec}\,x\cot x$, not the harder standalone $\int\text{cosec}\,x\,dx$ result. It is named here only to complete this document's own inventory of where spec 2.1's function definitions surface elsewhere on the paper; the full mark-scheme detail and the correction itself live in that other document.

(v) The confirmed boundary — what a WMA13 paper has never required, stated by a real examiner report rather than assumed. codex-deep-dive-differentiation-rules.md (§4) already documents this and is not reworked here: the Oct 2020 Q8(ii) examiner report names a third solution route some candidates attempted, rewriting an equation as $y=\arcsin(e^x)$ and differentiating arcsin directly, and states plainly that this route is beyond the specification. Spec 2.1 requires knowing $\arcsin$, $\arccos$, $\arctan$'s definitions, graphs and domains — never their derivatives — and this is the one place in the archive where a real Pearson document says so outright rather than leaving the boundary to be inferred from the spec's own silence.

#5. Ledger of every real exam anchor gathered for this specific topic

#6. What remains open — reported, not smoothed over

• The standalone-anchor count is exactly one. SIGNAL's own frequency table is explicit that this spec item is embedded in most series rather than separately tallied; this document does not know, and does not guess, how many of those embedded appearances would look like Jan 2022 Q2 if fully transcribed, because they have not been.
• No real WMA13 question asking directly for the domain or range of $\arcsin$, $\arccos$ or $\arctan$ by name was located in any source reviewed. The two "domain of the inverse left unstated" citations the lesson already carries are about inverse functions generally (spec 1.2); §3.D above restates that caveat rather than letting the citation imply more than it supports.
• Oct 2020 Q8(i)'s further marks, for the stationary-point stage that follows the differentiation, were not transcribed in any source available to this pass — the 2-mark differentiation ladder above is real and sourced; anything past it is not invented to complete the picture.
• No likelihood, percentage, or grade-boundary claim is made anywhere in this document. Every frequency statement above is either "the one located anchor," a real mark count from a real question, or a direct quote of SIGNAL's own "embedded, not separately tallied" finding — never a probability.

#Sources

• research/veridian/WMA13-verified-facts.md — §1 (spec 2.1–2.2 verbatim), §3 (formula-booklet/memorise split), §4 (mark-scheme conventions), §5 (the full 2.1–2.2 write-up, the Jan 2022 Q2 re-extraction, the Jan 2025 Q1(a) and Oct 2020 Q8 entries), §9 (this lesson's own 2026-09-01 audit-fix log).
• content/veridian/curriculum/wma13/reciprocal-and-inverse-trigonometric-functions.ts — the real lesson, read in full to confirm every definition, derivation, worked example, trap and MCQ it already contains.
• content/veridian/technique-hub/wma13/codex-formula-and-method-bank.md, forge-exam-technique-guide.md, signal-frequency-and-boundaries.md — read/grepped in full for this topic's existing coverage and frequency figures.
• content/veridian/technique-hub/wma13/codex-deep-dive-differentiation-rules.md (§4, the Oct 2020 Q8(ii) arcsin-beyond-spec finding) and codex-deep-dive-integration-by-recognition.md (§3.5, the Jun 2023 Q9(c) correction) — cited by name, not reworked.
• content/veridian/question-bank/wma13.ts — grepped; confirmed to contain no reciprocal- or inverse-trig item.