⚠ The single most useful thing to know about this chapter is that the four thousand and the five thousand are told with different vocabulary, deliberately, and that the difference is what makes sense of the quiz at vv19–21. The twelve baskets left over from the five loaves (6:43, already on these pages) are kophinoi — the small, sturdy wicker hand-basket a Jewish traveller carried his own food in, mocked by Roman satirists as standard Jewish kit. The seven left over here are spyrides — a large provisions hamper with no ethnic marking at all, the kind of thing you could lower a grown man over a city wall in, which is exactly what the book of Acts later does with one (Acts 9:25, not yet on these pages). Searched across the whole Greek New Testament, the two words never once trade places. Kophinos stands in six verses and every one of them is the five thousand and its twelve — in Matthew, in Mark, in Luke and in John, four writers telling one meal. Spyris stands in five, and four of them are the four thousand and its seven; the fifth is the Damascus wall. Two feedings, two containers, and not one slip anywhere in the New Testament. This translation keeps them apart in English too: hand-baskets at v19, hampers at v8 and v20.

⚠ “I am moved with pity” (v2) is one of exactly two places in the New Testament where Jesus says this verb of himself in the first person — and the other is Matthew’s telling of this same meal (Matthew 15:32, already on these pages), in the identical words. Everywhere else the verb is narrated about him from outside: at the feeding of the five thousand, Mark reports that he was moved for a crowd like sheep without a shepherd (6:34). Splanchnizomai is built on splanchna, the inward organs — the ancient world located strong feeling in the gut rather than the heart, so this is not a decorous sentiment but something felt in the body, and the sentence that follows it is about people giving out on a road.

Two different words of grace are said over this one meal, two verses apart, and this translation keeps them distinct: Jesus gave thanks over the loaves (v6, eucharistēsas) and blessed the fish (v7, eulogēsas). At the first feeding it was the blessing-word over the bread (6:41, already on these pages); at this Gospel’s Last Supper it will be the blessing-word over the bread and the thanksgiving-word over the cup (14:22–23, not yet on these pages). The vocabulary shifts around; the fourfold gesture — took, blessed or gave thanks, broke, gave — does not.

⚠ Dalmanutha (v10) is named once in the New Testament and nowhere else in it, and no one knows where it was. Matthew’s account of the same boat trip lands instead at “Magadan” (Matthew 15:39, already on these pages) — also a name that occurs exactly once in the New Testament and never again. Two Gospels, one crossing, two place-names that each appear exactly once in the whole New Testament and have never been securely identified on the ground. Copyists found this as awkward as modern readers do. At Matthew 15:39 the repair is visible in the apparatus this translation works from: a later branch of manuscripts replaces the unknown Magadan with Magdala, a town that certainly existed. Mark’s Dalmanutha is reported to have attracted the same kind of substitution, though that one does not appear in the apparatus this translation reads. Either way, this translation prints what its source text prints and does not tidy the geography.

The chapter opens “in those days,” with no interval given — the preceding chapter left Jesus in the Gentile Decapolis (7:31, already on these pages) and the narrative has not moved him since, which is the whole basis for the common reading that this second crowd is a Gentile one. Mark himself never says so. The reading is an inference from where the story left him standing, not a statement the text makes; the only crowd detail Mark gives is that some had come a long way.

⚠ Verse 12 does not finish, and this translation does not finish it for him. The Greek reads, word for word, “Truly I say to you, IF a sign will be given to this generation” — and then stops. It is not a damaged text and not a question. It is a Hebrew oath, carried into Greek intact: a self-curse whose second half is left unspoken because saying it aloud would be to call the penalty down. The full form runs may God do such and such to me, IF I do this thing — so the surviving “if” clause alone functions as the strongest possible denial. The same construction stands behind God’s own oath at Psalm 95:11, where the Hebrew is literally “im yevo’un — IF they shall enter into my rest” and the sense everyone reads out of it is “they shall NOT enter” (not yet on these pages). What Jesus means is flatly plain: no sign is coming. What he actually says is a broken sentence, and the break is the force of it.

All thirteen witnesses on the two shelves smooth it into an ordinary negative, which is a defensible thing to do with an idiom and also loses the whole texture. KJV and ASV read, word for word, ‘There shall no sign be given unto this generation’; GNV reorders the same words, ‘a sign shall not be given unto this generation’, and DRB reads ‘a sign shall not be given to this generation’; NIV has ‘no sign will be given to it’; NWT 1984 ‘No sign will be given to this generation’; and in Spanish RV 1909 reads «no se dará señal á esta generación» beside RV60’s «no se dará señal a esta generación». TLB abandons the sentence altogether for ‘Certainly not. How many more miracles do you people need?’, which is a paraphrase of the meaning with nothing of the wording left. ⭐ One small witness to how the Greek was read: NWT 1984 prints ‘Truly I say’ with no ‘to you’ — following an early critical edition that omitted the pronoun, a genuine split in the manuscripts recorded in the apparatus this translation works from.

⚠ Matthew’s version of this refusal keeps an exception; Mark’s has none. At Matthew 16:4 (already on these pages) the answer is ‘no sign will be given it EXCEPT the sign of Jonah’. Mark leaves the refusal absolute, with nothing held back for later. This is a real difference between the two Gospels, not a translator’s choice, and it is worth noticing which way it runs: the shorter, harder form is Mark’s.

“Groaning up from his spirit” (v12) translates anastenaxas, a compound that occurs exactly once in the whole New Testament — here. The simple verb means to groan; the prefix means up, the sound coming from depth. One chapter earlier he “sighed” over a deaf man with a different, simpler word (7:34, already on these pages). Mark reaches for the heavier compound only once, and it is not for a suffering man but for a demand for proof.

⚠ “The leaven of Herod” (v15) is Mark’s alone; the other Gospels name a different second party each time. Matthew’s telling of this same warning reads ‘the leaven of the Pharisees and Sadducees’ (Matthew 16:6, already on these pages, and repeated at 16:11–12, where Matthew adds his own explanation that the leaven means their TEACHING). Luke has ‘the leaven of the Pharisees, which is hypocrisy’ (Luke 12:1, not yet on these pages). Three Gospels, three different things paired with the Pharisees — a rival religious party, a political dynasty, and a named vice — and only Mark leaves the figure unexplained. He does not say what the leaven stands for, and the disciples, who are busy counting loaves, do not ask.

Leaven in this vocabulary is not a neutral word for yeast. It is the small hidden thing that changes the whole lump, which is why Paul can say ‘a little leaven leavens the whole batch’ twice, in two letters (1 Corinthians 5:6 and Galatians 5:9, neither yet on these pages), and why the Passover requires its removal from the house entirely.

⚠ “He kept ordering them” (v15) is the same verb this Gospel has used three times already, and every one of the others is a command to keep quiet. Jairus and his wife are ordered to tell no one (5:43); the Decapolis crowd is ordered to tell no one, twice in one verse (7:36, both already on these pages, and both discussed there). Here the same verb, now in the tense of repeated action, is spent on a warning rather than a silencing — and it is the one instance where the thing ordered is not obedience but attention.

⚠ “Is your heart hardened?” (v17) is the third time Mark reaches for this root, and the second time he turns it on the Twelve. The word is medical before it is moral — pōros is a callus, the bony matter that forms over a healed break, tissue that has stopped being able to feel. Mark used it first of the opponents watching a withered hand in a synagogue (3:5), then of the disciples after the five thousand (6:52, both already on these pages, where this Gospel’s own dictionary entry promised its return here). Those are the three, and the diagnosis Mark first gave the men plotting to kill Jesus he has now given the men following him, twice, without softening it either time.

⚠ “Having eyes, do you not see? And having ears, do you not hear?” (v18) is prophetic language, quoted without attribution. The nearest wording is Jeremiah’s indictment of a foolish people — ‘eyes they have and they do not see, ears they have and they do not hear’ (Jeremiah 5:21, not yet on these pages) — and Ezekiel says nearly the same to a rebellious house (Ezekiel 12:2, not yet on these pages). Mark names neither. He has done this before, in the chapter just past, reaching for a rare word out of Isaiah at exactly the healing that fulfils it and never citing the prophet (7:32, already on these pages). And the charge lands on the disciples in language this Gospel has already used for OUTSIDERS: ‘so that seeing they may see and not perceive, and hearing they may hear and not understand’ (4:12, already on these pages).

The quiz itself (vv19–20) is the reason this chapter needs its two basket-words kept apart — see the opening note. Jesus asks two questions about two different meals, and the disciples answer both correctly, twelve and seven, with the right container word for each. They have the data and not the point, which is precisely what verse 21 says. ⚠ And the question is left unanswered. Matthew closes the identical scene with ‘then they understood’ (Matthew 16:12, already on these pages). Mark writes “Do you not yet understand?” and moves straight on to a blind man.

⚠ This is the one healing in Mark that does not work the first time, and neither Matthew nor Luke has the scene at all. Matthew goes straight from the leaven argument to Caesarea Philippi with nothing between them (already on these pages, 16:12 to 16:13). Luke skips this whole stretch of Mark’s narrative entirely — his account runs from the twelve baskets of the five thousand directly to ‘who do the crowds say that I am’ (Luke 9:17–18, not yet on these pages), with the four thousand, the leaven, Dalmanutha and this blind man all absent. Mark reports a first touch that produces partial sight and a second that completes it, and offers no explanation for the two stages.

⚠ The man’s own sentence is broken Greek, and this translation leaves it broken: “I see people — because I see them like trees, walking.” Two different verbs for seeing in eight words, a causal “because” that does not do the work a “because” should, and a participle dangling on the end. Whatever else it is, it is what a man says when his sight is half back and he is trying to describe something he has no practice describing. Eight of the thirteen witnesses tidy it away; five keep the causal joint that makes it strange. KJV compresses to ‘I see men as trees, walking’, DRB to ‘I see men as it were trees, walking’, NIV to ‘I see people; they look like trees walking around’, and TLB expands it into a whole speech — ‘I see men! But I can’t see them very clearly; they look like tree trunks walking around!’ — adding a sentence the Greek does not have. The five that keep the “because” are ASV, ‘I see men; for I behold them as trees, walking’, NWT 1984, ‘I see men, because I observe what seem to be trees, but they are walking about’, the Spanish RV 1909, «Veo los hombres, pues veo que andan como árboles», TNM 1987, «porque observo lo que parece árboles», and — with the simile inverted, which is a slip of its own — GNV, ‘I see men: for I see them walking like trees’.

⚠ “Clearly, right into the distance” (v25) translates tēlaugōs, a word used exactly once in the New Testament and built out of tēle, far (the same element in ‘telescope’), and augē, radiance — far-shining, and so of sight, seeing at a distance and seeing sharply, both at once. The shelf splits on which half to keep. ASV takes the clarity alone, ‘saw all things clearly’; so does NIV, ‘he saw everything clearly’, and NWT 1984, ‘seeing everything distinctly’. GNV takes the distance, ‘saw every man afar off clearly’, and so does the Spanish shelf — RV 1909 reads «vió de lejos y claramente» and RV60 «vio de lejos y claramente» — four centuries apart, in two languages, the same decision.

⚠ One letter in verse 25 decides whether he sees everything or everyone, and the shelf divides exactly along the manuscript line — four of the thirteen witnesses one way, nine the other. The critical text reads hapanta, a neuter plural: he saw all THINGS. A later branch reads hapantas, masculine: he saw all MEN, every person. KJV follows the later reading, ‘saw every man clearly’; so does GNV, ‘saw every man afar off clearly’, and the Spanish RV 1909, «á todos», and RV60, «a todos». Those are the four; the other nine follow the earlier reading, among them ASV, ‘saw all things clearly’, and NVI, «ver todo con claridad». The same branch of manuscripts supplies the second half of verse 26, and the same four carry it: where this translation’s source text has only “Do not even go into the village,” KJV adds ‘nor tell it to any in the town’, GNV adds ‘nor tell it to any in the town’, RV 1909 «ni lo digas á nadie en la aldea» and RV60 «ni lo digas a nadie en la aldea». DRB carries a third form of its own, ‘Go into thy house, and if thou enter into the town, tell nobody’. The remaining eight print the bare command — ASV ‘Do not even enter into the village’, NWT 1984 ‘But do not enter into the village’, NVI «No vayas a entrar en el pueblo». It is a small addition and an understandable one — it turns an odd instruction (don’t go home through town) into the familiar one this Gospel gives everywhere else (don’t tell anyone).

A small vocabulary detail worth keeping: Mark writes that Jesus spat into the man’s ommata (v23), a rarer, more elevated word for eyes than the ordinary ophthalmoi he uses two verses later. The only other place that word appears in the New Testament is Matthew 20:34 (already on these pages), where Jesus touches the ommata of two blind men outside Jericho. Both of its New Testament occurrences are a hand on a blind man’s eyes.

Caesarea Philippi (v27) sits about forty kilometres north of the Sea of Galilee, at the foot of Mount Hermon by one of the springs that feed the Jordan — a city built around an older shrine to the god Pan, renamed for Augustus by Herod’s son Philip, and carrying Caesar’s name twice over by the time this conversation happens in its outlying villages. It is as far north as this Gospel takes Jesus inland; the only place it takes him further north at all is Sidon, on the coast, one chapter ago (7:31, already on these pages). The encyclopedia entry has the full history. Mark does not comment on the setting; he simply notes where the road was.

⚠ Peter’s confession here is four Greek words long, and everything that has been built on this scene for two thousand years is in the other Gospel. Mark: ‘You are the Christ’ — and then a rebuke into silence. Matthew 16:16–19 (already on these pages) has Peter add ‘the Son of the living God’, and then Jesus answer with the blessing on Simon Bar-Jonah, the name-play on Peter and the rock, the promise to build his congregation, and the keys of the kingdom. Mark has none of it — not the blessing, not the rock, not the keys. A Gospel that many readers have taken to stand especially close to Peter’s own memory is the Gospel that omits Peter’s commendation and keeps his rebuke. This translation reports the difference and does not attempt to explain it; the explanations on offer belong to different traditions and none of them is the text.

The guesses (v28) are all guesses about a man come back — John the Baptist, whom Herod has just beheaded in this same Gospel (6:27, already on these pages) and whom Herod himself feared had been raised (6:16); Elijah, who by his own story never died and so needs no raising; or one of the prophets. Mark’s list runs to three items; Matthew’s runs to four, with the name Jeremiah inserted before the catch-all (Matthew 16:14, already on these pages). And the question is put differently in the two Gospels: Mark’s Jesus asks “who do people say that I am,” while Matthew’s asks who they say ‘the Son of Man’ is — Matthew putting the title into the question that Mark keeps for the answer.

⚠ “He rebuked them” (v30) is not a mild word. It is the verb this Gospel uses when Jesus silences an unclean spirit — ‘be muzzled, and come out of him’ (1:25) — and when he silences the sea (4:39, both already on these pages, and the identical muzzling verb in both). All thirteen witnesses soften it here to charging, warning or ordering, and not one prints a word as strong as the Greek’s own: KJV and ASV both read ‘charged them that they should tell no man of him’, GNV ‘sharply charged them’, DRB ‘strictly charged them’, NWT 1984 ‘strictly charged them’, NIV ‘warned them not to tell anyone about him’, RV 1909 «les apercibió», RV60 «les mandó», NVI «ordenó». This translation keeps “rebuked”, because the same verb is about to be used twice more in three verses, of Peter and of Jesus, and the scene is unreadable if the three are not visibly the same word.

⚠ Three rebukes in four verses, all the same Greek verb, and the middle one runs the wrong way. Jesus rebukes the disciples into silence (v30); Peter rebukes Jesus (v32); Jesus rebukes Peter (v33). Mark does not remark on it. A disciple using his teacher’s own exorcism word against him is the whole distance the chapter has travelled from “you are the Christ” to “Satan” in four verses.

⚠ “Get behind me” (v33) and “come behind me” (v34) are the same two Greek words, and this translation prints them the same way even though the second is strange English. Opisō mou is what a follower does: it is the phrase behind ‘Come after me, and I will make you fishers of men’ (1:17), and the phrase John the Baptist uses of the one coming after him (1:7, both already on these pages, where this translation reads it the ordinary way, “after me,” because nothing in either verse turns on the picture). Here everything turns on the picture, so it is printed literally twice. So “get behind me” is not primarily a dismissal — it is a relocation. Peter has got in front, telling Jesus where the road goes; he is being sent back to the place a disciple occupies, which is behind. The very next sentence spells out the terms of standing there. Matthew’s parallel puts the same two sentences back to back in the same order (Matthew 16:23–24, already on these pages), with one addition Mark lacks — ‘you are a stumbling block to me’ — and one subtraction: Matthew has Jesus say the following words to the disciples, while Mark calls over the crowd as well.

⚠ “After three days” (v31), not “on the third day.” Mark writes meta treis hēmeras in all three of his passion predictions — here, and again at 9:31 and 10:34 (neither yet on these pages). Matthew and Luke both write ‘on the third day’ in the corresponding places (Matthew 16:21, 17:23 and 20:19, all already on these pages; Luke 9:22, not yet on these pages). Counted strictly the two are not the same span, and Matthew himself seems aware of it: the only time his Gospel uses Mark’s phrase, he puts it in the mouths of the chief priests quoting Jesus back at Pilate (Matthew 27:63, already on these pages). The shelf mostly follows whichever text it is translating, but GNV goes its own way here and reads ‘within three days rise again’ — a third span, neither after nor on. KJV, ASV, DRB, NIV and NWT 1984 all keep ‘after’ or ‘three days later’; so do RV 1909 and RV60, «después de tres días».

“He was speaking the word out in the open” (v32, parrēsia) is the note Mark chooses to add. This Gospel has spent eight chapters on silencings — demons muzzled, healed men told to say nothing, a confession rebuked into secrecy three verses ago. The first thing said plainly, with nothing held back, is that he is going to be killed.

⚠ Verses 35, 36 and 37 all turn on the same Greek word, psychē, and seven of the thirteen witnesses change words in the middle of the argument. KJV reads ‘whosoever will save his LIFE shall lose it’ at v35 and then ‘gain the whole world, and lose his own SOUL’ at v36 — the identical noun, two English words, one verse apart. NIV, GNV and DRB all make the same switch, and so do the Spanish RV 1909 and RV60, «vida» at v35 and «alma» at v36. ⭐ The other six hold one word across all three verses and let the argument stand as one argument: ASV keeps ‘life’ throughout, NWT 1984 keeps ‘soul’ throughout, and NVI keeps «vida», as do the NWT’s own 2013 revision and both editions of the Spanish TNM. This translation follows the ASV’s reading, ‘life’, and keeps life in all three, because the paradox in v35 is unintelligible with ‘soul’ (a soul is not the thing a person saves by holding on to it) and v36’s exchange is a merchant’s question about the same stake. The word’s genuine double range — the animating self, and the life that self is living — is real and is not resolved by this choice; the dictionary entry has it in full. What this translation refuses to do is let a reader see two different things where the Greek says one, because the sentence only works if the thing you could save in v35 is the thing you cannot buy back in v37.

⚠ “Take up his cross” (v34) is the first time the word appears in this Gospel, and every other time it appears, it is the real one. Searched across Mark, stauros and its verb stand here and then nowhere again until the passion and the empty tomb, chapters 15 and 16, where the crowd shouts ‘crucify him’, Simon of Cyrene is conscripted to carry the actual beam, and the passers-by tell him to come down from it. Mark has just said, three verses earlier, that the Son of Man will be KILLED — he has not said how. The image in v34 is the reader’s first sight of the method, arriving as a figure of speech before it arrives as a fact.

⚠ A flagged rendering. NWT 1984 prints ‘torture stake’ here rather than ‘cross’, and the Spanish TNM «madero de tormento» — a doctrinal rendering, held on the grounds that the Greek noun in older usage means an upright stake and that the shape of the Roman instrument is not stated in the text. Both halves of that are true about the word’s history; what the word denoted in the first century, under Roman execution practice, is the disputed part. Every other version on both shelves reads ‘cross’ or «cruz». The larger point the English loses either way is the one Cicero made about the penalty itself, which the dictionary entry carries: this was not a symbol to a first-century hearer, and a sentence telling a crowd to pick one up and carry it was not a metaphor they had to reach for.

“Disown himself” (v34, aparnēsasthō) is worth marking for later: Mark uses it three more times, and all three belong to one scene — Peter in a courtyard (14:30, 14:31 and 14:72, none yet on these pages). The man who has just been told to disown himself is the man who will disown someone else instead, and Mark gives the word to him here first.

The chapter ends on shame (v38) — an ordinary social word in a shame-and-honour culture, not a private feeling. Being ashamed of someone meant declining to be publicly associated with them, and the sentence answers exactly in kind. It is also where an earlier scene quietly returns: the generation that demanded a sign at v12 is the generation named again at v38, “adulterous and sinful,” and the sign it is finally given is the one Peter has just tried to talk Jesus out of.