On Food and Cooking: The Science and Lore of the Kitchen

This is a reference book, not an experiment. However, the book systematically tests and explains:

**Chemical transformations** during cooking: Maillard reactions, caramelisation, protein denaturation, starch gelatinisation, emulsification, and fermentation.

**Ingredient composition**: what proteins, fats, carbohydrates, water, and minerals are in common foods (meats, vegetables, fruits, grains, dairy, eggs, seafood, spices) and how their molecular structure determines behaviour during cooking.

**Historical and cultural cooking methods**: how different societies have solved the same culinary problems (e.g., tenderising meat, preserving vegetables, leavening bread) and what the underlying science reveals about which methods are most effective.

**Sensory science**: the chemical basis of flavour, aroma, texture, and colour — including how taste receptors work, volatile aroma compounds, and the role of temperature and pH in perception.

**Food safety and spoilage**: microbial growth conditions, enzyme activity, and the science behind preservation techniques (salting, curing, fermenting, freezing, canning).

The "comparator" is implicit: traditional cooking lore versus scientific explanation. The "outcome measures" are mechanistic understanding and predictive power — can you explain *why* a technique works, and can you predict what will happen if you change a variable?

No human subjects were studied in a controlled trial. The "subjects" are:

**Food ingredients**: hundreds of specific varieties of meats, vegetables, fruits, grains, dairy products, eggs, seafood, fungi, spices, and processed foods.

**Cooking processes**: dozens of techniques (roasting, braising, frying, steaming, poaching, grilling, baking, fermenting, curing, pickling, smoking, sous-vide, pressure cooking, microwave cooking).

**Historical and contemporary sources**: the book draws on peer-reviewed food science literature (primarily from the 1950s–2000s), agricultural research, culinary texts, and ethnographic accounts of traditional food preparation across cultures.

The "setting" is the laboratory kitchen and the scholarly library — not a clinical trial environment.

McGee does not report original measurements. Instead, he synthesises measurements from thousands of primary research papers. Key measurement methods referenced throughout the book include:

**Chemical analysis**: gas chromatography-mass spectrometry (GC-MS) for volatile aroma compounds; high-performance liquid chromatography (HPLC) for sugars, amino acids, and vitamins; spectrophotometry for colour and pigment analysis; pH meters for acidity; water activity meters for microbial growth prediction.

**Physical measurements**: texture analysers for tenderness and crispness; differential scanning calorimetry for protein denaturation temperatures; rheometers for viscosity and gel strength; thermocouples and infrared thermometers for internal cooking temperatures.

**Sensory evaluation**: trained taste panels using descriptive analysis (scoring specific attributes like sweetness, bitterness, umami, astringency on 0–100 scales); triangle tests for difference detection; hedonic scales for preference.

**Microbiological assays**: plate counts for bacterial load; challenge studies for pathogen survival under different cooking and storage conditions.

The book does not provide individual study effect sizes or p-values because it is a synthesis, not a meta-analysis. Instead, it reports consensus findings and ranges (e.g., "meat is considered medium-rare at an internal temperature of 54–57°C"; "most vegetables lose 30–50% of their vitamin C during boiling").

### Study design

This is a **narrative synthesis** — a comprehensive, non-systematic review of the scientific literature on food and cooking, combined with historical and cultural analysis. It is not a systematic review, meta-analysis, or randomised controlled trial. The author, Harold McGee, is a writer and food scientist who spent decades reading primary research, conducting his own kitchen experiments, and consulting with chefs and food scientists.

### How information was gathered

McGee does not specify a formal search strategy (no PRISMA checklist, no explicit inclusion/exclusion criteria). He draws on:

Peer-reviewed journals in food science, chemistry, biology, and nutrition (e.g., *Journal of Agricultural and Food Chemistry*, *Journal of Food Science*, *Cereal Chemistry*, *Meat Science*)

Classic textbooks and monographs (e.g., *The Chemistry of Cooking* by Belle Lowe, *Food Chemistry* by Belitz and Grosch)

Historical cookbooks and agricultural treatises (e.g., Apicius, Brillat-Savarin, Escoffier)

Personal correspondence and collaboration with chefs (e.g., Heston Blumenthal, Ferran Adrià) and food scientists

### What this design can and cannot prove

**What it can do:**

Provide mechanistic explanations for why cooking techniques work at a molecular level

Identify general principles that apply across many foods (e.g., "heat denatures proteins, causing them to coagulate and firm up")

Reveal contradictions between traditional lore and scientific evidence (e.g., "searing meat does not 'seal in juices' — it actually causes greater moisture loss")

Offer a framework for designing personal experiments by identifying key variables (temperature, time, pH, water activity, enzyme activity)

**What it cannot do:**

Provide precise effect sizes or confidence intervals for specific interventions

Establish causal relationships in a controlled, replicable manner

Account for individual variation in human taste perception, metabolism, or health outcomes

Replace primary research for questions about specific health effects (e.g., "does eating charred meat cause cancer?" — the book discusses the chemistry of heterocyclic amines but does not conduct epidemiological analysis)

### Major methodological weaknesses

**No systematic search strategy**: studies were selected based on the author's judgment, introducing potential selection bias

**No quantitative synthesis**: effect sizes are not pooled, so you cannot determine the average magnitude of an effect across studies

**Outdated in places**: the second edition was published in 2004; some specific claims about nutrition (e.g., dietary fat and heart disease) reflect the science of that era and have been superseded

**Anecdotal evidence**: some claims are supported by kitchen experiments or chef testimony rather than peer-reviewed research

**No conflict of interest disclosure**: the book is a commercial publication; McGee has consulted for the food industry, but specific financial relationships are not disclosed

Because this is a synthesis, findings are organised by topic rather than as a single result. Below are representative, well-supported conclusions from the book:

### Meat and poultry

**Collagen breakdown**: Tough cuts of meat become tender when cooked to an internal temperature of 60–70°C for extended periods (hours), because collagen (a connective tissue protein) hydrolyses into gelatin at these temperatures. Above 70°C, collagen shrinks and squeezes out moisture, making meat drier.

**Myoglobin and doneness**: The red colour in meat comes from myoglobin. At 54–57°C (medium-rare), myoglobin denatures partially, turning pink. At 60–65°C (medium), it turns brown. At 71°C (well-done), it is fully denatured and grey. This is why internal temperature, not cooking time, determines doneness.

**Searing and moisture**: Contrary to popular belief, searing does not "seal in juices." In controlled experiments, seared meat loses 15–25% more moisture during subsequent cooking than unseared meat. The browning (Maillard reaction) creates flavour compounds, not a moisture barrier.

### Vegetables

**Cell wall structure**: Plant cells are held together by pectin, a polysaccharide that dissolves in acidic conditions. Adding acid (vinegar, lemon juice) to cooking water helps vegetables retain firmness by stabilising pectin. Adding alkali (baking soda) breaks down pectin, making vegetables mushy — but also speeds cooking and preserves colour in green vegetables.

**Green vegetable colour**: Chlorophyll degrades in acidic conditions, turning olive-brown. To preserve bright green colour, cook green vegetables in boiling water (which drives off volatile acids) for the shortest time possible, then shock in ice water. Adding baking soda preserves colour but destroys thiamine (vitamin B1) and makes texture mushy.

**Nutrient loss**: Boiling vegetables in large volumes of water leaches 30–50% of water-soluble vitamins (C, B vitamins) into the cooking water. Steaming or microwaving with minimal water reduces losses to 10–20%.

### Eggs

**Coagulation temperatures**: Egg white proteins begin to coagulate at 62°C and are fully set at 70°C. Egg yolk proteins coagulate at 65°C and are fully set at 70°C. This narrow window (62–70°C) is why precise temperature control matters for poached, soft-boiled, and sous-vide eggs.

**Overcooking**: Heating eggs above 80°C causes the proteins to over-coagulate and squeeze out water, resulting in a rubbery texture and greenish-grey ring around the yolk (caused by iron from the yolk reacting with sulphur from the white to form ferrous sulphide).

### Bread and baking

**Gluten development**: Gluten is formed when two proteins (glutenin and gliadin) in wheat flour combine with water and are worked (kneaded). The network traps gas produced by yeast or chemical leaveners, causing the dough to rise. Over-kneading breaks the gluten network, while under-kneading leaves it too weak to hold gas.

**Yeast activity**: Yeast produces CO₂ most rapidly at 27–32°C. Above 38°C, yeast dies. Below 10°C, yeast becomes dormant. This is why proofing temperature matters — too hot kills the yeast, too cold slows it down.

**Starch gelatinisation**: During baking, starch granules absorb water and swell at 60–70°C, thickening the dough and setting the crumb structure. This is why bread is fully baked when the internal temperature reaches 90–95°C — the starch has fully gelatinised.

### Flavour and aroma

**Maillard reaction**: This reaction between amino acids and reducing sugars produces hundreds of volatile aroma compounds and brown pigments. It begins at around 140°C (for dry heat) and accelerates with temperature. This is why roasting, grilling, and frying produce more complex flavours than boiling or steaming.

**Salt and perception**: Salt suppresses bitterness and enhances sweetness and umami at low concentrations (0.5–1.0% by weight). This is why a pinch of salt improves the flavour of coffee, chocolate, and tomatoes — it reduces perceived bitterness and amplifies other flavours.

Because this is a synthesis, "effect magnitude" is reported as ranges and typical values rather than single numbers:

**Collagen tenderisation**: Tough meat (e.g., beef chuck) cooked at 65°C for 4–6 hours becomes approximately 50–70% more tender (measured by shear force) compared to cooking at 100°C for 1 hour. The difference is the difference between "chewy" and "fall-apart."

**Vitamin C loss**: Boiling broccoli for 10 minutes in a large volume of water results in 40–50% loss of vitamin C. Steaming for 5 minutes results in 10–15% loss. The practical effect: a serving of boiled broccoli provides roughly half the vitamin C of the same weight of raw broccoli.

**Egg coagulation**: A soft-boiled egg cooked at 100°C for 4 minutes has a white that is just set (62–65°C internal) and a yolk that is still liquid (55–60°C internal). Cooking for 6 minutes at 100°C raises the yolk to 65–68°C, producing a jammy yolk. Cooking for 10 minutes at 100°C raises the yolk above 70°C, producing a fully set, crumbly yolk. The difference of 2 minutes changes the texture from "runny" to "custard-like."

**Maillard reaction intensity**: Roasting a chicken at 180°C produces moderate browning after 45 minutes. Roasting at 220°C produces the same degree of browning in 25 minutes — but also produces higher levels of acrylamide (a potential carcinogen) and heterocyclic amines. The trade-off is flavour intensity versus potential health risk.

### Author-acknowledged limitations

**Incompleteness**: McGee explicitly states that the book cannot cover every food or technique in exhaustive detail, and that readers should consult primary sources for specific questions.

**Evolving science**: He notes that food science is a rapidly advancing field and that some conclusions may be revised as new research emerges.

**Cultural bias**: The book focuses primarily on Western cooking traditions, with less depth on Asian, African, and Indigenous cuisines.

### Critical reader observations

**No systematic review methodology**: Without explicit search criteria, inclusion/exclusion rules, or quality assessment of studies, the book cannot claim to be comprehensive or unbiased. Important studies may have been missed or selectively cited.

**Outdated nutrition science**: The 2004 edition predates major shifts in understanding of dietary fat, cholesterol, gut microbiome, and ultra-processed foods. For example, the book's discussion of saturated fat and heart disease reflects the lipid hypothesis of the 1990s, which has been substantially revised.

**Anecdotal evidence**: Some claims (e.g., about specific cooking techniques used by famous chefs) are supported only by personal observation or hearsay, not controlled experiments.

**No quantitative synthesis**: Readers cannot determine the statistical reliability of any claim. When McGee says "most vegetables lose 30–50% of vitamin C during boiling," this is a range from multiple studies, but we don't know the number of studies, the variability, or the quality of the evidence.

**Potential conflicts of interest**: McGee has consulted for food companies and restaurants. While the book is generally critical of food industry practices (e.g., industrial processing, additives), there is no formal disclosure of financial relationships.

**Single-author perspective**: Unlike a textbook with multiple expert contributors, this is one person's synthesis. Errors or omissions in one area (e.g., fermentation science, molecular gastronomy) may not be caught by peer review.

For someone running their own n=1 experiment in the kitchen:

### What to test

**Temperature precision**: Test the effect of cooking a steak to different internal temperatures (50°C, 55°C, 60°C, 65°C, 70°C) using a sous-vide circulator or probe thermometer. Measure tenderness (by chewing), juiciness (by weighing before and after cooking), and flavour (by blind taste test).

**Acid and vegetable texture**: Cook green beans in plain boiling water vs. water with 1 tablespoon vinegar per litre vs. water with 1/2 teaspoon baking soda per litre. Measure cooking time to desired tenderness and rate colour (1–5 scale, 5 = bright green).

**Salt and bitterness**: Brew a cup of black coffee. Add 0.5g salt (a small pinch) to one cup and leave another unsalted. Taste blind and rate bitterness (1–10 scale, 10 = most bitter). Repeat with grapefruit juice, dark chocolate, or arugula.

**Egg cooking precision**: Cook eggs at 63°C (sous-vide) for 45 minutes, 65°C for 45 minutes, and 70°C for 45 minutes. Compare texture of white and yolk. Rate on a 1–5 scale for creaminess, firmness, and overall preference.

### Minimum meaningful duration

**Single-session tests**: Most cooking experiments can be completed in one cooking session (30 minutes to 4 hours). For example, testing different steak temperatures requires one afternoon.

**Repeated tests**: For reliable results, repeat each condition at least 3 times on different days with